GLP Hot Topic- Bioanalytical Incurred Sample Re-Analysis

GLP Hot Topic- Bioanalytical Incurred Sample Re-Analysis

Scott C. Rumsey, RQAP-GLP

GLP Quality Consulting, LLC

During the last several years the FDA has focused attention upon the accuracy of Bioanalytical sample analysis. Specifically, questions have arisen about the accuracy of results reported from previously validated methods when large discrepancies are found in reanalysis of samples. These “incurred” or study samples, frequently had results which varied beyond the QC performance criteria.

Until recently, the FDA’s expectations for Bioanalytical method validation was consistent with the 2001 Guidance Document entitled “Guidance for Industry: Bioanalytical Method Validation1,” where precision and accuracy were determined through the use of spiked QC’s and standards during the validation process and verified in each analytical sample analysis Run. This guidance document was largely based upon an FDA/Industry conference report “Analytical Methods Validation: A revisit with a Decade of Progress.2” However, recent FDA inspections and presentations at industry forums, such as the Society of Quality Assurance Annual meeting and at the 3rd AAPS/FDA Bioanalytical Workshop demonstrate a shift in expectations. A summary of these current expectations for incurred sample analysis have been published in an AAPS Journal articles entitled “Workshop/Conference Report—Quantitative Bioanalytical Methods Validation and Implementation: Best Practices for Chromatographic and Ligand Binding Assays3” and “Confirmatory Reanalysis of Incurred Bioanalytical Samples.4”

Incurred samples may differ from spiked QC samples for a variety of reasons. Small molecule samples may have metabolites present which are not included in spiked QC’s. These metabolites may revert to the parent molecule and result in non-reproducible results. In macromolecules, background or matrix interference may occur which was not encountered in method validation. Protein-binding differences in patient samples, recovery issues or non-homogeneity of samples may also affect results.

The 3rd AAPS/FDA Workshop included recommendations for incurred sample evaluation for each species in GLP studies. This evaluation was only needed once for each species (i.e. it is not necessary to repeat the analysis for multiple studies in the same species). Additionally, it was recommended that the results of the incurred analysis reproducibility testing be reported in the final Bioanalytical report and/or as an addendum to the method validation report. The consensus was that study samples used for incurred analysis reproducibility may be used for comparison purposes, but do not necessarily need to be used for calculating reported sample concentrations. Any observed lack of accuracy (unless it is determined that it is due to analyte instability or inter-conversion) should have an investigation performed to determine the reason and the impact upon the study.

Following the meeting, there were many discussions about various aspects of incurred sample analysis. For example, (1) how many samples would need to be analyzed? (2) What range of concentration(s) would need to be analyzed? (3) What would be an acceptable range of difference between reanalyzed samples? (4) What will be the FDA approach to enforcement of incurred sample analysis? These questions and others were discussed in another industry meeting at the AAPS Workshop on Current Topics in GLP Bioanalysis: Assay Reproducibility for Incurred Samples—Implications of Crystal City Recommendations held February 7-8, 2008.

The following points are recommended to comply with current FDA GLP expectations/industry consensus:

●For GLP studies, analyze incurred samples from at least one species/matrix/method. Guidance from the 2007 AAPS Workshop would be followed for pre-clinical studies.

●A written Incurred sample analysis SOP should be developed to describe: frequency, selection of samples, acceptance criteria, reporting and investigation of unexpected results.

●There is a preference for utilizing specific study samples over pooled samples.

●The incurred sample analysis should be conducted early in sample analysis, taking into account laboratory efficiency.

●Use the same number of replicates as was used in the study sample analysis (i.e. single, duplicate, etc.)

●The number of samples used for incurred sample analysis must be meaningful such that conclusions about assay reproducibility can be made. It was suggested that it was preferable to analyze samples from multiple subjects, rather than a complete profile from a few subjects, to better account for subject to subject variability. A sampling time point at or near cmax and another in the elimination phase per subject was recommended.

●Acceptance Criteria: A fixed limit approach was preferred to the statistical approach. It was suggested that 2/3 of the samples be within 20% of the initial values for small molecules and 30% for large molecules. The criteria should be stated in the SOP, and must be scientifically justifiable.

●Results of incurred sample analysis can be reported either as an amendment to the method validation report or included in the study results. FDA stated that it would be preferable to include the results in the study report, as method validation reports are often not included in submissions.

●An investigation must be conducted for any incurred sample analysis failure. The investigation should be documented and the effect on the study assessed.

C.T. Viswanathan of FDA stated that the agency expects everyone to perform incurred sample analysis. Enforcement will be progressive and escalate. For future inspections you should have an SOP/study plan to describe incurred sample analysis. He said that current guidance requires accurate and reliable study sample results. Enforcement, therefore, is consistent with that requirement. In the future, the FDA may revise the 2001 Bioanalytical Guidance Document to specifically include incurred sample analysis. He stated that it is not necessary to retrospectively conduct analysis for incurred samples. However, if non-reproducibility is seen in submissions and incurred sample analysis was not performed, the agency may reject the study.

1FDA/CDER Guidance for Industry: Bioanalytical Method Validation. Rockville, MD: US Department of Health and Human Services FDA (CDER) and (CVM); May 2001.

2 Shah VP, Midha KK, Findlay JW, et.al. Bioanalytical Method Validation—A revisit with a decade of progress.Pharm Res.2000, 17:1551-1557.

3Viswanathan CT, Bansal S, Booth B, et. al. Workshop/Conference report—Quantitative Bioanalytical Methods Validation and Implementation: Best Practices for Chromatographic and Ligand Binding Assays. AAPS Journal. 2007, 9E30-42.

4Rocci ML, Devanarayan V, Haughey DB, Jardieu P. Confirmatory Reanalysis of Incurred Bioanalytical Samples. AAPS Journal. 2007, 9E336-343.

Scott Rumsey can be contacted at (269) 558-8115 or at