Monitoring stability chambers. The CAL must monitor the stability chambers' performance on a daily basis. Chart recorders and electronic probes can monitor temperature and humidity. Chart recorders should be monitored every day. They are vulnerable to malfunction when pens run out of ink and other basic errors occur. Electronic probes combined with a commercial environmental-monitoring program constantly evaluate the chambers and provide performance-trending data, so that malfunctions may be predicted before they occur. If a chamber deviates from its setpoint (e.g., it exceeds ±2 °C or 5% RH), the system can be programmed to call an employee to investigate and resolve the problem. If the CAL uses chart recorders, it may not detect the error until the chamber is checked during the daily routine. The sponsor should review the CAL's procedure to determine whether the process is appropriate and adequate. If electronic probes are used, their performance should be confirmed periodically (e.g., every six months).

Sample storage. If the sponsor is storing the stability samples in its own chambers or in a vendor's chambers, it must decide on the timing for pulling samples and shipping them to the CAL. During large studies, the CAL may have a team waiting to receive and test the samples. If a shipment is delayed, it might compromise the CAL's ability to test the samples within the test-window period.

If the CAL is to store the stability samples in its chambers, then the parties should decide in advance whether the CAL or the sponsor will decide how the samples are distributed and how much overage will be placed on station. A recommended overage is a minimum of 25% additional samples in case the study is extended or additional samples are required. The party responsible for labeling the samples should be specified clearly. This decision will help the CAL plan and place the samples on station. Once the samples are on station, the pull dates must be communicated to the sponsor. For large, complex studies, the sponsor may choose to audit the stability setup.

Study design

The US Food and Drug Administration accepts lot and packaging matrices for stability studies that are based on ICH guidelines. The guidelines greatly reduce the number of samples to be tested but increase a study design's complexity. Special tests such as in-use tests, which may be triggered by the failure of the corresponding open-dish tests, also contribute to the complexity.

Studies often have contingency conditions for when failures are observed. For example, samples may be stored at long-term conditions of 30 °C and 65% RH and at 25 °C and 60% RH. The latter samples might only be tested if the former samples fail specifications. Ideally the former samples' results should be approved by the quality assurance unit before the latter samples are tested. The two sets of samples should therefore be placed on station at staggered intervals to allow the first set to be evaluated to avoid erroneously testing the contingency samples.

The anticipated time of the beginning of a study should be clearly communicated and mutually decided by both parties. When a CAL commits to a study, its resource allocation is based on the information that the sponsor provides for the start date. The CAL considers resource allocations to determine whether they may conflict with current active studies. If delays such as manufacturing difficulties occur, then the CAL should be notified to ensure that no resource conflicts with the new proposed start date. Open communication also serves the interest of the sponsor because it may require the CAL to report the analyses quickly to meet a filing deadline. Such deadlines should be communicated and mutually agreed upon in advance.

Summary

If the CAL's procedures are not consistent with the sponsor's expectations, then specific details may be addressed in the quality assurance agreement or the study protocol. In addition, the CAL and the sponsor's lead technical staff should meet to evaluate the study. This evaluation should include a comparison of the instrumentation specified in the methods and the CAL's equipment. Items such as the compatibility of available HPLCs may be addressed (e.g., dwell volume comparison), as may questions such as whether a wrist-action or reciprocating shaker is preferred.

The sponsor should evaluate the CAL's departmental structure to determine the necessary resources. The sponsor should be informed about whether the same person will remain the technical lead throughout the study and whether the same resources will be used for each time point. The sponsor should also learn whether temporary employees will be used for large stability pulls, which occur during the first six months and include accelerated conditions and special time points such as photostability. The early time points generate critical data, and the sponsor must be confident that the teams are sufficiently trained. Inexperienced temporary employees can reduce the study's quality.

Conclusion

A sponsor that seeks to hire a CAL to perform stability studies should consider several important factors and take critical steps during its evaluation process. The sponsor should perform a thorough audit of the CAL's procedures, paying attention to the latter's training program, environmental chambers, data processing, and reporting practices and timelines. Timing and study design should be considered so that the appropriate resources are allocated and the sponsor receives analytical reports in a timely manner. The relationship between the sponsor and the CAL should be viewed as a partnership with open communication between both parties. The parties should have the same objective: to perform the stability study within the expected timelines so that the product can get to market successfully.

David C. Browne is a manager of stability and pharmaceutical testing at Intertek QTI, 291 Route 22 East, Whitehouse, NJ, 08888, tel. 908.534.4445, ext. 658, fax 908.534.1054, david.browne@intertek.com

Reference

1. ICH, Q1A Stability Testing of New Drug Substances and Products (Geneva, February 2003).