Analytical Test Method Validation for API Raw Material, In Process Control, and Early Intermediate Material Tests
Introduction
This procedure provides guidance for the validation of analytical test methods. These analytical test methods include those tests which evaluate API Raw Materials, In Process samples (e.g. reaction monitoring), and early intermediate materials (prior to the introduction of the first critical intermediate).
Per ICH Q7A, the degree of analytical validation performed should reflect the purpose of the analysis and the stage of the API production process.
The evaluation of robustness should be considered during the development phase and depends on the type of procedure under study. If measurements are susceptible to variations in analytical conditions, the analytical conditions should be suitably controlled or a precautionary statement should be included in the procedure. One consequence of the evaluation of robustness may be that a series of system suitability parameters (e.g., resolution test) is established to ensure that the validity of the analytical procedure is maintained whenever used. Critical parameters effecting response factors, if used in the method, should be identified and characterized during robustness testing.
Recommendations and Rationale
Method robustness is established when the effect of various parameters is evaluated with respect to the requirements of the method. The experiments performed to demonstrate robustness of the method may be expanded during the validation exercise to further support the rationale to reduced SST. If possible, validate with worst-case scenarios (e.g. when developing a method for varying grades of material (crude and pure).
Solution Stability Experiments
It is recommended that sites perform solution stability experiments. The results should be analyzed for signs of degradation to determine if the solution is stable for the period that is studied.
The following example approaches may be considered. The acceptance criteria should be based on the limits’ range. The extent of change should not significantly affect the final result. The change should also not affect the decisions made from the data.
Approach 1:
As directed by the test method, prepare standard and sample aliquots and analyze them. The test samples are allowed to stand, under normal conditions of test (e.g., at room temperature), for a minimum length of time equivalent to the maximum expected use time, (typically 24 hours to one week). Sample and/or standard stability are demonstrated for more than 24 hours if applicable. If possible, analyte stability is demonstrated over a time period that slightly exceeds the stability time period indicated in the test method. During this study, solutions are analyzed against freshly prepared solutions. For acceptance a minimum discernible trend in analyte response from initial and final analyses is observed and analyses should agree within reproducibility found for the system precision.
Approach 2:
For standard stability for a low level impurity method, two different stock preparations of equal concentration are prepared (a1 and b1) and diluted separately to the same solution concentration (a2 and b2). Six (6) injections of standard check solution “a2” and three (3) injections of standard check solution “b2” are performed. From each set of injections calculate the mean peak area response for the analyte main peak then calculate the standard check using the following equation.
Check = Mean Area STD “a2” x Concentration STD “b2”(µg/ml) x 100
Mean Area Std “b2” x Concentration Std “a2”(µg/ml)
Approximately 50ml of standard check solution A is decanted into a flask clearly labeled and stored in a refrigerator (+2 to +8 degrees C). The remaining volume is stored at room temperature. A fresh standard check solution is prepared on the day of analysis and the standard check procedure is repeated for each of the stored standard A solutions against the freshly prepared check solution after a period of 24, 48, 72 hours and 7 days storage. For acceptance criteria, the standard check is between 95% and 105% (any acceptance criteria applied must consider the concentration of the standard solutions under test, for example the acceptance range may vary from a 10ppm solution (0.001%) to a 0.1% solution). Standard stability may be performed over a longer period if necessary.
Approach 3:
For a chiral HPLC method, solution stability is assessed using an injection and analysis of the sample of the appropriate test material at the following times after preparation.
- 0 hours (i.e., within 1 hour of preparation)
- 24 hours
- 48 hours
Use one of the samples prepared for the precision or robustness studies for the sample stability study, i.e., the first sample injected for the precision study may be used to obtain the t = 0 data-point. Extra time points may be added and some of the tests may fail the expected criteria. However, this is recommended to be explained in the protocol ahead of time. These extra time points are examined to determine at which point the solutions are no longer stable.
Approach 4:
For TLC where the sample is required to be analyzed immediately, the standard only is analyzed and the intensity should be the same as at t=0 and the plate should not have new spots.
Approach 5:
For an HPLC method where the standard peak is a retention time marker only, the criteria for solution stability is for the main peak to be present and no new ones eluted. Repeatability studies cover the range of typical sample preparation time therefore sample solution stability is combined with repeatability studies.
Approach 6:
For a HPLC Limits test, both samples and standard are examined. The criteria for the sample is based on the repeatability study but the criteria is more of a Pass/Fail scenario rather than numerical. In this case the standard is retention time marker and criteria are that the same peaks are present and no new ones eluted.
Approach 7:
For some non stable (intermediate) compounds it is also possible to verify solution stability by comparing detector response versus time.
Recommended Robustness Data:
Data obtained from studies for robustness are usually documented as part of the method development report. A reference to these studies along with a summary should also be given in the validation report. If not documented in the method development report, robustness studies should be documented in the validation report.
In some instances a set of experiments may be able to satisfy several validation parameters i.e. robustness, intermediate precision and repeatability could be combined and studied together instead of as isolated effects.
- For example, for compendial physical tests such as LOD/ROI/pH where the analyst and sample amount are primary contributors to variation and the method is well defined in general chapters are ideal for this situation.
- An experiment could be constructed such that variations in sample size, analyst and equipment could all be studied at once.
- High, medium and low typical sample amounts may be analyzed once each by two analysts using two different pieces of equipment.
- The composite RSD would be analyzed and compared to the acceptance criteria.
- Here repeatability is addressed across different typical sample amounts, intermediate precision is addressed by different analysts and equipment and robustness is shown across typically allowed variations in the method.
- If the method is shown to be reliable across all of these variations, each factor alone does not need to be demonstrated.
Recommended Robustness Criteria
Changes within the test range whether allowed explicitly or implicitly by the assay should not exceed the previously defined validation parameters for accuracy, precision or specificity. This may be accomplished by evaluation of system suitability parameters that are relevant to the change.
For example, robustness for limits tests should confirm that variations still cause a pass/fail decision to be unaffected by the change. Multiple preparations below and above the specification can be used to demonstrate the ability of the method to reliably distinguish passing and failing results.
Increased robustness testing during development may provide additional support for an abbreviated System Suitability Testing (SST). If robustness testing is not adequately performed and documented during development and/ or validation, there is more of a reliance on detailed SST, which should be included during the run. A full SST at the beginning of a campaign could be performed, and then repeated periodically throughout the campaign. As a working practice, some sites allow 24 hours validity between SST and running a sample (provided no major changes in instrument operating conditions have been performed within the time period). This allows the laboratory to analyze a series of samples within that 24 hour period without repeating the SST and provides the advantage of allowing a quicker sample turn around in cases where analysis of many samples may be required over a short time period (e.g. when monitoring residual solvent by hourly sampling during drying). The choice of taking this approach should be carefully weighed with the risk of implicating a large amount of data if a system suitability failure were to occur.