Incurred sample reproducibility

The topic of incurred sample reanalysis (ISR) was introduced at the Crystal City III Conference in May 2006 to help understand the poor reproducibility of results found by FDA in some cases when samples from studies were reanalyzed. The recommendation from the workshop report [10] on this topic is that an adequate evaluation of incurred sample reproducibility should be conducted for each species used for GLP toxicology experiments and in selected clinical studies. It is also recognized that the degree of reproducibility could be different in human samples in comparison to the animal samples. Selection of the studies to be evaluated for ISR was left up to the sponsor. 

In addition to analytical and nonanalytical repeats, some samples in certain studies need to be reassayed to demonstrate incurred sample reproducibility (ISR). This type of testing is a critical step for demonstrating the reproducibility of the bioanalytical method with samples from dosed subjects (as distinct from precision demonstrated with QCs, i.e., blank matrix spiked with drug). Written procedures in this regard should include a description of how the samples are selected for reanalysis, the comparison and reporting of the original and repeat results, and the acceptance criteria for variability between results. 

Additional Components Analyte recovery (extraction efficiency) Stock stability Matrix stability Post process stability (extract) Integrity of dilution Carryover and contamination Matrix effects Incurred sample reproducibility Incurred sample stability... 

Method reproducibility — Individual incurred samples from four subjects (approximately 5% of all samples) were re-assayed individually to evaluate reproducibility. The four samples set for reanalysis and evenly spaced throughout the study were designated 101, 123, 145, and 166. The values generated from the reassays were used only to assess reproducibility and were not used in pharmacokinetic calculations. Table 2.2 summarizes the method reproducibility results. The analytical method used in study M06-830 was accurate, precise, and reproducible. 

Fig. 12 (a, top) High internal standard responses were observed for incurred samples only. Analyte repaglinide extraction automatic liquid-liquid extraction, (b, middle) Postelution infusion results show that ion suppression existed near the retention time of the analyte (1.57 min) from the pooled control blank used for the preparation of calibration standards and quality controls, (c, bottom) Absence of ion suppression near the retention time of the analyte in subject predose sample. Reproduced from ref. [36] with permission from Elsevier... 

Fig. 19 Randomly scattered low internal standard (IS) responses observed for incurred samples only, whose IS responses were within normal range during repeat analyses. Analyte olanzapine IS olanzapine-d3 sample pretreatment at clinic 25 % (w/v) L-ascorbic add added to plasma in a ratio of 1.25 100 (v/v) extraction MCX (mixed-mode strong cation exchange)-based solid-phase extraction. An incurred sample was coded for reassay when its IS response was outside 50 % of the mean IS response of the accepted calibration standards and quality controls. Reproduced from ref. [36] with permission from Elsevier...

Precision and accuracy With method established perform three runs Perform at least six runs Apply appropriate statistics Evaluate reproducibility of incurred samples in each toxicology species Clearly defined and acceptable Evaluate reproducibility of incurred samples in key studies... 

An additional bioanalytical parameter of great current interest [54] is reproducibility of analysis of samples from dosed subjects (the so-called incurred samples). Whether such analyses are to be conducted in all species or most appropriately... 

Given the foregoing discussion of some of the unique characteristics of macromolecules that lead to clear differences in their pharmacokinetics compared to those typical of small-molecule drugs, there is a subset of the entire group of bioanalytical assay validation parameters that are of key importance in support of pharmacokinetics of candidate macromolecular therapeutics. Assuming demonstration of accuracy and precision of sufficient quality for the intended application of the assay (e.g., non-GLP discovery support or GLP toxicokinetic support, as discussed above), the most important characteristics of a given assay in support of pharmacokinetic studies are likely to be selectivity, specificity, and reproducibility for analysis of incurred samples. These are all related to the ability of the LBA to detect and quantitate solely, or as closely as possible to solely, the analyte of interest. 

Fast, D.M., Kelley, M., Viswanathan, CT., O Shaughnessy, J., King, S.P., Chaudhary, A., Weiner, R., DeStefano, A.J., and Tang, D. (2009) Workshop report and follow up AAPS workshop on current topics in GLP bioanalysis assay reproducibility for incurred samples implications of Crystal City recommendations. The AAPS Journal, 11,238 241. 

The repeat results from this type of testing are used to confirm the reproducibility of the assay and should be presented in the bioanalytical study report as a separate data table. In cases where repeat testing of incurred samples does not confirm that the method is reproducible, the bioanalytical laboratory should investigate the cause of the nonreproducibility before continuing its use of the method or reporting results from samples previously assayed. 

Although there are many issues to consider in the conduct of appropriate and relevant ISR assessments, particular attention should be paid to the number of samples selected for reassay. To provide adequate coverage across a study in its entirety, 5 10% of the total sample size should be reassayed, with the 5% minimum limited to larger studies. It is also important to note that acceptance limits for ISR comparisons should be commensurate with the methodology (e.g., chromatographic vs. ligand binding assays) and demonstrated assay performance wide acceptance limits are not recommended. Further details for these issues and other considerations pertinent to ISR can be found in the 2009 workshop report and follow-up publication from the February 2008 AAPS Workshop on assay reproducibility for incurred samples [8]. 

Number of laboratories Number of incurred samples Number of dose levels Recovery Reproducibility... 

The fundamental components of a bioanal5dical validation (and aU validations in general) are selectivity sensitivity accuracy precision range of reliable response and linearity and reproducibility. Other characteristics of the method that should be addressed are stability carryover and control of lab contamination and matrix effects, including interferences from metabolites in incurred samples or other co-extracted compounds, as well as ionization suppression. 

The most recent bioanalytical Workshop Report (Viswanathan 2007) devotes considerable space to this topic and some recommendations not discussed previously (Section 9.4.7b) are included below. There should be some assessment of both reproducibility and accuracy of the reported concentration. Sufficient data should be generated to demonstrate that the current matrix (i.e. the incurred sample matrix) produces results similar to those previously validated. It is recognized that accuracy of the result generated from incurred samples can be more difficult to assess. It requires evaluation of any additional factors besides reproducibility upon storage, which could perturb the reported concentration. These could include metabolites converted to parent during sample preparation or LC-MS/MS analysis, matrix effects from high concentrations of metabolites, or variable recovery between analyte and internal standard (Viswanathan 2007). Most of these phenomena are those described previously (Jemal 2002) and discussed in Section 9.4.7b. 

Figure 11.35 Diagram illustrating a method for providing absolute quantitation of metabolites in incurred serum samples when no analytical standard is available. The method exploits the trend towards equimolar responses for a drug candidate and its metabolite at low flowrates ( 10 nL.min ) that permits measurement of the concentration ratio of metabolite to parent compound via the observed ratio of signals for the two using nanospray ionization (1 nano/Ip,nano)- However, use of a fast nonselective extraction procedure like protein precipitation for the LC-ESl-MS/MS assays is not suitable for the nanospray experiment since the extracts are too complex, so a more selective procedure, liquid-liquid extraction (LLE) in this case, must be used to prepare these extracts. The connection between the two procedures is achieved by analyzing the LLE extracts by both nanospray and the LC-ESI-MS method used for the assays of the incurred samples, yielding an LC-ESI-MS peak area ratio (A] lc/ms/- p,lc/ms) comparison with (iM.mmo/Ip.nano) that is interpreted as the concentration ratio. This comparison yields a calibration factor k ji = (lM,nano/Ip,nano)/(AM.LC/Ms/Ap Lc/Ms) that enables quantitation of the metabolite relative to that of the parent compound subsequently, absolute quantitation of the latter by conventional LC-MS using an analytical standard to prepare a calibration curve permits absolute quantitation of the metabolite also. Only a limited number of nanospray analyses (one in the example shown) is required to derive a k j value applicable to a complete set of study samples. Reproduced from Valaskovic, Rapid Commun. Mass Spectrom. 20, 1087 (2006), with permission of John Wiley Sons Ltd.

The previous assay was applied to the analysis of SDZ in salmon tissue, with some modifications. The re-extraction with phosphoric acid solution was replaced by SPE on an SCX cartridge preconditioned with MeCN and phosphoric acid solution. The cartridge was washed with MeCN, and SDZ was eluted with an MeCN phosphoric acid mixture. The eluate was injected directly into the chromatographic system, followed by postcolumn derivatization under similar conditions to the previous assay. The derivatization time was 1.2 min, and the fluorescence intensity was approximately a quarter of that for optimal conditions. However, the postcolumn derivatization was found to be considerably less labor intensive and was easily reproducible (recoveries 83-85% CV < 7%). A significant improvement in the LOD value was obtained (0.2 yug/kg) (160). The SDZ residues from incurred salmon tissue were confirmed by MS detection however, the sample cleanup should be improved due to the lack of sensitivity of MS. Therefore, SDZ residues were eluted from the SCX SPE cartridge with phosphoric acid, and the eluate was concentrated on a C18 SPE cartridge preconditioned with MeOH and water. The residues were eluted with MeCN, and the eluate was evaporated to dryness and reconstituted prior to the analysis. The column effluent was delivered into the atmospheric-pressure ion source, and SIM was chosen for positive ions at m/z 251, 158, 156, and 108, respectively (161). 

Surfaces of disperse materials are less well defined than that of mercury. Problems are Incurred with heterogeneity, non-zero solubility, non-inertness on titration, the presence of impurities remaining after the preparation, etc. Some uncertainty as to the specific surface area Ag also presents itself even if a reproducible value is determined when using an established method such as BET (Nj) gas adsorption some reservation remains about its physical meaning do Ions adsorbing from solution "see" the same surface as Nj molecules adsorbing from the gas phase at subzero temperatures on evacuated samples Because of... 

A two-tier approach is often utilized by residue control laboratories whereby samples are first screened to identify the suspected positive (non-compliant) samples, which are subject to further quantitative and confirmatory analysis. Screening methods should be inexpensive and rapid and permit a high sample throughput. The basic criteria that should be met are a detection capability below the RL, a low incidence of false-negative (compliant) results (<5%), and a high degree of repeatability, reproducibility, and robustness. A low incidence of false-positive results is also important to reduce the costs incurred by additional confirmatory analysis. False-positive results in screening analysis can occur for a number of reasons, such as if the test is sensitive to other structurally related compounds naturally present in the matrix or to co-contaminants. 

Figure 6.11 UPLC QqTOF MS chromatograms of a tylosrn (4.1 p.g/kg, RSD = 1.5%, n = 3) incurred honey sample (CE—collision energy). Plots A and C traces of tylosin A. Plots B and D traces of tylosin B. Plots A2—spectrum at 4.42 min from plot A1 Plot B2—spectrum at 3.88 min from Bl plot C2—spectrum at 4.43 min from plot Cl. Plot D2—spectrum at 3.87 min from plot D1. Proposed fragmentation is based on the nitrogen rule and accurate mass measurement. (Reproduced and modified from Wang and Leung ° with permission from the Canadian Food Inspection Agency published by John Wiley Sons copyright 2007. Crown in the right of Canada.)...

---