Assay, variations

Factors such as assay variations, age, and prostate gland size are known to affect cutoff values. Also, free to total PSA cutoffs are influenced by the sensitivity and specificity values chosen, the reflex range for total PSA used, differences in free PSA assays, and differences in populations studied. Different PSA values are considered due to differences in cutoffs in different assays. Studies have shown that the comparison of a chemiluminescent free PSA showed a 25 percent difference in values. These types of variations suggest a need for standardization (9,29). 

The effect was minor at about 87% of signal strength compared to control hybridization to probes without spikes. This may be within the range of assay variation. However, what is particularly attractive about the dye-labeled dNTP route is that it is relatively inexpensive compared to primer labeling. Moreover, spot characterization (e.g., spot morphology, signal intensity, spot diameter) could be easily determined from the scans (Figure 4.40). 

The hydrogel coatings achieved the lowest limits of detection (LTD) 1300 to 1600 amoles/spot, but exhibited significant assay variation (22% intra-slide to 37% inter-slide CV). LLD levels of surface-modified polysfyrene slides (Maxisorb black. Nunc) equaled 1500 amole/spof af 15 to 32% CV, while reflective (mirror-like) slides coated with 3-aminopropyltriethoxysi-lane (Amersham) showed the lowest variahon with CVs at 11 to 14%. 

In a protocol about collaborative studies [10] it is also considered what is called preliminary estimates of precision. Among these the protocol defines the total within-laboratory standard deviation . This includes both the within-run or intra-assay variation (= repeatability) and the between-run or inter-assay variation. The latter means that one has measured on different days and preferably has used different calibration curves. It can be considered as a within-laboratory reproducibility. These estimates can be determined prior to an interlaboratory method performance study. The total within-laboratory standard deviation may be estimated fi-om ruggedness trials [10]. 

The intra-assay variation for plasma was performed in a tenfold analysis and gave CVs of 0.8-4.1% with the exception of tryptophan, methionine, and the imino acids (6.7-8.9%). Urine behaved similarly to plasma in the sense of reproducibility, with CVs between 3.6 and 8.6% in 23 analyses over a 6-month period. Exceptions in this respect were the amino acids with a low ninhydrin color yield, such as / -alanine and sarcosine. 

The intra- and inter-assay variation is determined in tenfold analyses of a pool plasma sample. Table 3.4.2 shows the results of these measurements. The linearity of this method should be assessed for all analytes. Pristanic acid and the C26 0 fatty acid were linear up to 16 pmol/1, phytanic acid to 100 pmol/l and the C22 0 and C24 0 fatty acid to 200 pmol/1. The lower detection limit for all analytes was at a level of less than 0.01 pmol/l, for the lower reporting levels (LOQ), an analysis of a blank solution was taken into account. The blank levels of the analytes phytanic acid, pristanic acid, and fatty acids C22 0, C24 0 and C26 0 were 0.04, 0.01, 0.41, 0.68 and... 

Taylor and colleagues [98] at the Mayo Clinic published a method for the simultaneous analysis of urinary cortisol and cortisone. They used 2H4 cortisol as an internal standard and took a 0.5-ml urine sample. An API 2000 with Turboion-spray source was used in the positive-ion mode. Chromatography was conducted on a standard-bore C18 column with Q8 precolumn filter. MRM was conducted in the positive-ion mode monitoring m/z 363—>121 for cortisol, 367—>121 for 2TL, cortisol, and 361— -121 for cortisone. Cortisol and cortisone were separated and both were eluted within 2 min. Inter- and intra-assay variation for both compounds was < 9% for amounts above 2 pig/dl. The values obtained agree well with those of other studies, such as ours (Table 5.3.2) [62]. They found a range for cortisol for adult males of 4.2-60 pg/24 h and for adult females 3.0-43 pg/24 h. In summary, the 3-min run time of their method has allowed the Mayo group to completely transfer their cortisol and cortisone workload from RIA and HPLC to MS/MS. 

The objective of any population study is to detect and quantify variation. This objective can be accomplished in many ways depending on the overall thrust of a study. Often, the aim of a study is to follow the distribution of single polymorphisms.10,11 Assaying variation in population studies also may involve the characterization of levels of variation between individuals, within populations, between populations, or over zones of contact of populations. This objective requires methods for extensive screening of many individuals for polymorphisms. 

The analytical variation of the Cobas Integra ISE method in New Lab A was studied in the validation process by running the two-level system quality control samples forlO days, and Daytrol for8 days. The intra-assay variations (n = 20) of the two system controls were much better than those quoted by the manufacturer. The within-batch variation of the lower system control was verified as 0.43% compared to 0.81% given by the manufacturer. The higher level varied by 1.4% the manufacturer quoted a respective variation of 2.5%. 

Lab B reported its inter-assay variation as total variation of Daytrol only at the higher therapeutic level. Respectively, Lab C reported variation only at the lower therapeutic level (Table 1). As the national target coefficient of variation (CV%) for analytical variation is 2% for S-Li, all methods fulfilled this quality specification, when the long-term control, Daytrol, was used. Inter-assay variation of the lower level system control was 3.5 times higher, than quoted by the manufacturer. This stresses the point that every laboratory should check the performance specifications given by the manufacturer. The variation of the higher therapeutic level showed equality with the obtained validation results. 

Table 1 Inter-assay variation, CVmeas % of the three ion-selective electrode (ISE) applications during method validation and at the 6-month checkpoint for the lower and higher therapeutic levels and the toxic level of serum lithium (S-Li)...

We consider here a hypothetical hormone assay, for which the manufacturer or a research laboratory wants to estimate the LoD, The default values a = P = 5% are used. It is supposed that the manufacturer has 10 samples available from patients lacking the hormone because of disease or pharmacological suppression. Ten measurements of each blank sample are performed on 10 different days to ensure that the total assay variation is reflected. Only nonnegative values are provided by the assay, and the distribution of the 100 blank measurements is skew (Figure 14-5). Thus LoB is estimated nonparametrically as the 95th percentile of the measurement distribution. The 95th percentile corresponds to the 95.5 ordered observation (= 100 x (95/100) + 0.5). The 95 and 96 observations have the values 0.0539 and 0.0548 U/L, respectively, Linear interpolation between the 95 and 96 observations yields an LoB estimate of 0.0544 U/L (= 0.0539 + 0.5 X (0.0548 - 0.0539)). 

Brambilla D, Reichelderfer P, Bremer J, Shapiro DE, Hershow RC, Katzenstein DA, et al. The contribution of assay variation and biological variation to the total variability of plasma HIV-1 RNA measurements. AIDS 1999 13 2269-79. 

Reproducibility which applies to all 5 concepts, may refer to within-assay or between-assay variation. Sometimes within-assay reproducibility is designated as repeatability (Palmer and Cavallero, 1980)... 

SD itself does not yield useful information, unless compared to the mean (%CV Section 15.1). Between-assay variation is generally greater than within-assay variation, but care should be taken with the latter to reduce bias by randomly distributing the duplicates. The between-assay variation is usually of more value to estimate the accuracy of the procedure and, plotted on a quality-control chart, may indicate trends, such as deterioration of reagents (bias-type error Section 1.3). An example of the calculation of the within-assay and between-assay variability is given in Table 15.3. In this example, the SD of between-assay results is about 5-6 times higher than for the within-assay results. For a satisfactory EIA, this factor should be less than 2-3 and the between-assay variability should be less than 10%. 

The way a layout is used may change from validation to production assay. Only changes that the validation experiment can support with valid statistical inference will be permissible between validation and production use. Assuming there are replicate assays within each analyst and day, we can estimate the repeatability directly. Similarly, if there are replicates of a sample within each assay, we can estimate within assay variation in potency directly. These estimates of variation allow us to reliably predict the performance of the assay system with various numbers of replicates at each level where we have direct replication. For example, with two replicate samples at potency 1.0 in each assay, we can predict the precision of potency when one, two, three, or more within assay replicates are combined. Similarly, if there are replicate assays within analyst and day, we can use the variation among these replicates to predict the precision of... 

Rhodes A, Borthwick D, Sykes R, et al. The use of cell line standards to reduce HER-2/neu assay variation in multiple European cancer centers and the potential of automated image analysis to provide for more accurate cut points for predicting clinical response to trastuzumab. Am J Clin Pathol. 2004 122 51-60. 

Fig. 7 Calibration and residual plot for the inhibition assay-based detection of warfarin in plasma ultrafiltrate ( = 3). The mean normalized response value (RAG/RO) at each analyte concentration from three independent assays was used to calculate the calibration curve and to determine the assay variation [59]...

Takano, S., H. Komiyama, and M. Nomura. 2001. A study of the correction of assay variation caused by the reagents in the measurement of rat albumin. Japanese Journal of Clinical Chemistry 30 164—167. 

The third advantage of quantitative real-time RT-PCR is that it allows one to monitor the treatment effect in individual patients as a potential surrogate marker after adjuvant chemotherapy (K3) or immunotherapy (S4) against micrometastatic diseases. For the purpose of monitoring, a relative quantification method which is designed to determine exact, PCR efficiency-corrected mRNA concentration, normalized to a calibrator, might be desirable to overcome the inter-assay variation from run to run (SI3). This may make it possible to directly compare the mRNA values at different time points. 

Moller, P., Moller, L., Godschalk, R.W. and Jones, G.D. (2010). Assessment and reduction of comet assay variation in relation to DNA damage studies from the European Comet Assay Validation Group. Mutagenesis. Vol. 25 No. 2 pp. 109-111 ISSN 1464-... 

In the context of LOD, false positive and false-negative responses can be defined as follows false positive—a response for a blank that contains no analyte that falls above the LOD (i.e., a blank that is concluded to contain the analyte) false negative—a response for a sample that contains the analyte that falls below the LOD (therefore it is concluded that the sample does not contain the analyte at detectable levels). False-positive and false-negative responses are possible due to the assay variation (responses for a given sample are represented by a frequency distribution and not a single result). This scenario is illustrated in Fig. 8. To a certain extent, the definition and value for LOD... 

A separate analysis of the measured response from the reference samples (e.g., if a radial diffusion assay, the ring diameter or if an ELISA assay, the absorbance or if an HPLC assay, the peak height) will indicate the between-assay variation. The analysis of duphcate samples of the reference samples will indicate the within-assay variation. 

The reason for this discrepancy is not known. The fucosyltransferase studied by both groups was assayed in a similar manner. Grimes (1970) points out that there is a fucosidase in 3T3 and SV-40-3T3 cells which interferes with the enzyme assay variations in this fucosidase may be responsible for the different results reported by the two laboratories. 

Fig. 11. Narrow inter-assay and intra-assay variations are critical parameters for successful high-throughput screens using FLIPR technology. Some final advice...

The ID-GC-MS-SIM method of Adlecreutz and coworkers (1993) is perhaps the most frequently used in clinical and medical studies, although intra- and total-assay variabilities of all analytes are considered to be high, probably reflecting low-analyte concentrations in biological fluids (Wilkinson et al., 2002). Other GC-MS methods report lower inter-assay variation of 6%-ll% isoflavones in serum. The latest method of measuring urinary excretion of isoflavones (daidzein, genistein, equol. 

Immunoassay is one of the most sensitive analytical techniques described for phytoestrogen measurements in human biological fluids. Detection limits for target analytes are below 1 nmol/L for most methods with an average recovery of 92% and inter-/intra-assay variation less than 10% for most methods (WiUdnson et al., 2002). Immunoassays have been validated against other analytical techniques and excellent correlation has been demonstrated. Immunoassay can be a powerful alternative to GC-MS- or HPLC-based methods for studying the bioavaUabflity and metabolic fate of dietary isoflavonoids because of its ease, relatively low cost, availability for... 

---