Figure 3. Assessment Form for Test Method Suitability and Validation Status
Periodic method performance and compliance reviews (Figure 2, in green) are helpful, as is the assessment of the impact from
method modifications, to ensure an overall validation continuum. A recently published Food and Drug Administration (FDA) conference
presentation, given by Mary Malarkey of FDA's Center for Biologics Evaluation and Research, summarized FDA's current thinking
about production operations and a process validation continuum.9 Although this presentation predominately addressed only the process validation continuum, the concept of constantly monitoring
and adjusting (if needed) the manufacturing process is certainly similar to the proposed maintenance tools for analytical
methods herein. Acceptable method performance is continuously monitored with assay control charts, indicating that test systems
are in overall control. This evidence is usually submitted in annual reports to FDA. When a licensed biopharmaceutical drug
has been produced over several years and test methods have in the interim gone thorough inevitable changes, a more thorough
review may be needed to ensure that the production process is properly measured and under control. Many changes that are small,
and seemingly minor, could accumulate and compound into major changes in test method performance over the years. In addition,
regulatory expectations are changing and are generally becoming more stringent in their demands for accuracy, reliability,
and validity in test results. Test methods validated years ago may be suspect for their acceptable performance and current
validation status in an upcoming inspection. A thorough periodic review is therefore a good preventative tool. Rather than
being reactive to unacceptable method performance or regulatory inspection observations, periodic reviews may prevent unacceptable
method performance and inspection observations. Performance reviews should be triggered through the VMP as all AMV approval
dates should be listed within this plan. To ensure these reviews are done meaningfully and consistently among the various
test methods and laboratories, a review checklist is an excellent tool to periodically assess the validated state for each
test. This will ultimately ensure that each test method delivers accurate and reliable test results to support the quality
of the production process or product. It will also ensure a test method remains compliant, efficient, and economical because
the rate of invalid test results does not postpone production flow or product release, nor does it increase the demand of
expensive instrumentation, reagents, or personnel. A simplified checklist is shown in Figure 3.
Figure 4. Overlay of Control Charts for Process and Analytical Performance
Emergency method performance reviews are shown in red in Figure 2 to indicate the urgency and criticality of the associated
investigations. Emergency reviews are usually triggered by out-of-specification (OOS) results, above-action-level results,
or frequent invalid assay runs that may postpone the release of material. A good AMM will support the generation of accurate
and reliable test results, meaning here that the observed OOS results may, unfortunately, indicate a "bad" production batch
or that this OOS was statistically just one of those (expected) outliers. There is one more part of an AMM program that can
be used to assess process and method performance simultaneously, and with this, provides more confidence in all process results.
By overlapping each particular test result for each manufacturing process test point with the assay control results, one saves
resources for investigations because there is often an immediate answer for the validity of the assay result in question.
Provided that test system suitability was appropriately developed and maintained in the AMM program and the analytical test
system is under control, there is no need to investigate further. More detail on statistical process control and assay control
combination charts is provided in Figure 4.
