Info

biomarker, the test drug did not affect the assay at all test concentrations, which exceeds the highest expected concentration of the drug in the serum. The results illustrated that the drug affected only its target protein.

Stability In addition to sample integrity during collection, stability of protein biomarkers and assay reagents is required to show that the assay is not compromised by this preanalytical factor. Stability information is gathered during method validation on short-term storage and process stability, and is extended through in -study validation for long-term storage stability. Reagent storage stability is tested on reference material stock solutions and ligand- binding reagent stock solution if information from the supplier is lacking. Working solutions will be prepared from the stored stock solution and assay performance compared to those from that of a freshly reconstituted solution. Process stability is evaluated at different time durations and various conditions to mimic possible sample handling in the laboratory from the results of SCs subjected to stress from multiple freeze-thaw cycles and temperature exposures (e.g., four cycles and up to 16 to 72 hours to at 2 to 8°C and ambient temperature).

Long-term storage stability of the SCs should be started during validation and continued to support the drug development program. The acceptance criteria of storage stability are according to those established during method validation, prior to conducting the experiment. For long-term stability evaluation, in addition to comparison against the initial concentration, trend analysis can be more useful than the acceptance criteria strictly applied. This is due to the challenges of a kit lot change, potentially shifting the mean value of the SCs, and a new preparation of the standard curve, introducing systemic bias to add an error component that is not caused by the storage conditions (similar to that illustrated in Figure 4).

Assay Application

Before sample analysis begins, a written method standard operating procedure (SOP) must be in place. Often, a validation report should also be issued. Documentation of sample assay should follow the SOP in a GLP compliance manner (see the discussion later on regulatory issues), with traceable records. The data should be stored in a secure place for auditing.

Standard Operating Procedure A validated method should have defined procedures, assay performance parameters collected from experiments using the same method during prestudy validation, and acceptance criteria established reflecting the acceptability of the method performance. These elements constitute a method SOP and are discussed in the following section.

Method Procedures A written method procedure should clearly describe the following:

• The intended purpose of the method

• The principle of the method

• A validation performance summary

• Materials and reagents, including lot numbers and expiration dates

• Solutions preparation

• Standards and QC, SC preparations

• Sample pretreatment (processing)

• Analytical steps

• Data regression

• Acceptance criteria

• References

During the development of a novel biomarker, a series of method modifications can take place. It is important that specific names (version number) and effective dates be given to the same method as that used in a clinical study. A historical log should be kept on the modifications made for each version, and its application in specific studies identified. Sometimes, for various reasons, if a method change has to happen within a clinical study, a crossover comparison of the old and new methods applied to a subset of clinical samples must be designed with a statistical approach to show method equivalence.

Regression Model The response-concentration relationship of the LBA method is nonlinear. Curve fitting would require regression model selection based on multiple runs of standard curves from the method validation accuracy and precision experiments. Typically, a four- or five - parameter logistic model will fit most methods with a choice of weighting factors (such as 1/ response or 1/variance). The initial selection is based on the residual of the standard mean. It can be chosen to obtain the best fit over the entire range and with regard to the low end if sensitivity is an issue. The minimal total error of the VS data generated in the validation batches should be used to determine the regression algorithm and weighting. Once the regression model is chosen, it is a basic element of the method that should not be changed during assay implementation. Justification must be given and documented for a regression model change.

Acceptance Criteria Without sufficient data from the healthy and patient samples, the acceptance criteria for a novel biomarker can initially be set as determined by the assay performance from prestudy method validation. The biological data obtained from in-study validation from subject samples can then be used to refine the initial acceptance criteria set by the prestudy validation. The process of setting acceptance criteria for a protein biomarker measurement follows an evolving path. During the exploratory phase, the acceptance criteria were initially set according to the initial assay performance of prestudy method validation. After use in pilot studies, assessment should be made on the suitability of the assay stringency vs. the effect observed. The biological data obtained from the in- study validation from subject samples during in-study phase can then be used to refine the initial acceptance criteria set by the prestudy validation. For example, an assay with 50% total error may still be acceptable for detecting a twofold treatment effect observed in the in-study phase clinical trial. Setting the same acceptance criteria of a given method may be most convenient for an analytical laboratory. However, for a novel biomarker, it may not be the most appropriate for all of its applications. One could take into account the intended purpose of the application and the possible outcomes in the analytical phase. For example, the effect of one population or indication may be different from another (e.g., change from a twofold treatment effect into only 30%), which may require a more stringent method and/or more subjects to increase the predictive power in such an application for the other population.

Controls in Clinical Studies For a well-planned assay application, both analytical quality controls and biological sample controls should be available to assess the assay variability. Data of analytical variability are gathered from standards, QC and SC of each analytical run of sample analysis. A common set of SCs (or pools of incurred samples from a previous study) analyzed by multiple bioanalytical laboratories can provide an assessment of assay performance among the laboratories. Variability of supplies in reagents and reference standard materials can be detected by tracking the assay performance of a common set of SCs used within and between studies. An example is shown by an SC chart for serum C-terminal telopeptides of type I collagen (CTx) assay in Figure 6, using the Westgard rule for monitoring (Westgard, 2003) as well as a priori acceptance criteria of 25%. The low and high SCs were prepared by pooling a large volume of authentic samples, and the mean values were established during method validation. Figure 6 shows the Levey Jennings charts of more than 117 analytical runs from five clinical studies (A to F) over more than 2.5 years. During this time span, two reference stock materials and three kit batches were used. The plots show a slight positive bias in the SC values after the initial sequence of studies B and C. In addition, a slight negative bias was observed prior to study A. However, since the SC values were still acceptable within ±25%, no action was taken. Other statistical analysis can be use to assess SC data on other assay variability sources, such as operator differences and sample storage stability trend.

At the same time, data from predosed samples can furnish information on biological variability among the subjects and between studies. Figure 7 shows the baseline concentration distribution of TRACP 5b, a bone resorption bio-marker, in various patient populations from four clinical studies. Statistical tools can be used to construct distribution charts of the predosed samples of each population, and cohorts can be compared. For pilot clinical trials with small number of cohorts, one should note that the data may not be normally

Project Management Made Easy

Project Management Made Easy

What you need to know about… Project Management Made Easy! Project management consists of more than just a large building project and can encompass small projects as well. No matter what the size of your project, you need to have some sort of project management. How you manage your project has everything to do with its outcome.

Get My Free Ebook


Post a comment