Low limits of quantification

Bayliss and co-workers [10] combined ultra-high flow rates, parallel LC columns, a multiplex electrospray source, and mass spectrometric detection for the rapid determination of pharmaceuticals in plasma using four narrow bore (50 mm x 1 mm, 30 pm Oasis HLB) or capillary (50 mm x 0.18 mm, 25 pm Oasis HLB) HPLC columns with large particle sizes (to avoid high system back-pressure) in parallel with a multiple probe injector and a MUX MS interface. Small sample aliquots were injected directly into the system without sample pre-treatment procedure, obtaining very low limits of quantification (from 1 to 5 ng/mL). 

It permits fast (especially if coupled with recent UHPLC chromatographic system) and reproducible analyses with high selectivity (in MRM or SRM acquisition mode) to targeted compounds, together with low limit of quantifications (LOQ) - in the order of ng/mL or lower. 

In general, a satisfying method for quantitative analysis has to provide high sensitivity, low limits of detection (LOD), low limits of quantification (LOQ), a linearity range of at least three to four orders of magnitude, high precision (repeatability and reproducibility of data), and accuracy (experimental data as close as possible to the true value ). 

Polymeric microfluidic systems coupled to a microfabricated planar polymer tip can be used as a stable ion source for ESI-MS. A parylene tip at the end of the microchannel delivers fluid which easily produces a stable Taylor cone at the tip via an applied voltage. The described device appears to facilitate the formation of a stable spray current for the electrospray process and hence offers an attractive alternative to previously reported electrospray emitters. When this interface was employed for the quantification of methylphenidate in urine extracts via direct infusion MS analysis, this system demonstrated stable electrospray performance, good reproducibility, a wide linear dynamic range, a relatively low limit of quantification, good precision and accuracy, and negligible system carryover. We believe polymeric devices such as described in this report merit further investigation for chip-based sample analysis employing electrospray MS in the future. 

The linearity of the calibration curve was ascertained by adding known concentrations of manufactured lactose to cow milk over the concentration range of interest. Intraday precision was determined by analyzing 15 replicates of low- and high-quality control samples. Interday precision was determined by analyzing qnal-ity control samples over 15 consecutive days, and the coefficient of variation for lactose over two levels was calculated. The low limit of quantification was condncted by additional dilution until indication of peak-to-peak signal-to-noise ratio as 5 1 was achieved. To determine the recovery rate, lactose was added to cow milk samples. Absolute recovery was indicated by a ratio of the observed value to the corresponding expected value. 

Table 8.76 shows the main characteristics of voltammetry. Trace-element analysis by electrochemical methods is attractive due to the low limits of detection that can be achieved at relatively low cost. The advantage of using standard addition as a means of calibration and quantification is that matrix effects in the sample are taken into consideration. Analytical responses in voltammetry sometimes lack the predictability of techniques such as optical spectrometry, mostly because interactions at electrode/solution interfaces can be extremely complex. The role of the electrolyte and additional solutions in voltammetry are crucial. Many determinations are pH dependent, and the electrolyte can increase both the conductivity and selectivity of the solution. Voltammetry offers some advantages over atomic absorption. It allows the determination of an element under different oxidation states (e.g. Fe2+/Fe3+). 

There are special problems in bioequivalency determinations when conventional pharmacokinetic studies are not possible. For example, when drugs are administered intranasally for direct treatment of receptors in the nasal mucosa, the concentration of drug in plasma may be below the limit of quantification. In such cases we are forced to attempt measurement of clinical response. The subjectivity and/or low precision of this type of study can be a serious problem. 

Standards, controls, and samples (250 fiL each) were treated with 500 fiL acetonitrile-acetic acid (99 1 v/v) containing IS (2.50 jUg/mL), vortexed for 10 sec, incubated for 5 min, and centrifuged at 15,000 g for 5 min. The supernatants (1650 //L) were loaded onto a polypropylene 96-well plate containing 900 fxL HPLC water under low vacuum. The SPE plates were conditioned with 500 fxL methanol followed by 300 jx. acetonitrile-water-acetic acid (30 69.5 0.5 v/v/v) (solvent A), washed with 1000 /xL solvent A, dried under full vacuum for 10 min, wiped dry with paper, eluted with 500 jxL methanol-trifluoroacetic acid (99.9 0.1 v/v) (solvent B) and then with 400 //L solvent B for 2 min, evaporated to dryness at 65°C under a gentle air stream, reconstituted with 200 /xL methanol-hydrochloric acid (0.1 M) (70 30 v/v) and assayed. The injection volume was 50 i L. Figure 11.3 shows chromatograms of blank plasma and spiked plasma with lumefantrine. A calibration curve was constructed in a concentration range of 25 to 20,000 ng/mL. Intra-assay and interassay coefficients of variation were below 5.2 and 4.0%, respectively. The limit of detection was 10 ng/mL. The limit of quantification was 25 ng/mL. 

The use of robotics can be adopted also in sample preparation steps, in particular on-line SPE [7], This necessity is particular evident when small quantity of starting materials is available and the target molecules are present at low concentration levels. With the advent of miniaturization and automated procedures for samples handling, treatments and analysis, the lost of analytes due to a laboratory steps can be reduced. The reduction of analyte losses and the possibility to analyze even a total sample (no loss) leads to lower limits of detection (and consequently lower limits of quantification). Smaller volumes bring to obtain adequate sensitivity and selectivity for a large variety of compounds. In addition, on-line SPE requires low solvent consumption without the need to remove all residual water from cartridges, since elution solvents are compatible with the separation methods. 

Both the limit of detection and the limit of quantification have been also defined as ratios of the analyte signal to the background signal (S/N). Thus, an S/N ratio of 3 has been used to define the detection limit, whereas a S/N ratio of 10 has been used to define the limit of quantification. Determination of the signal-to-noise ratios is performed by comparing measured signals from samples with known low concentrations of analyte with those of blank samples. 

Both the detection limit and the limit of quantification, as defined, are often not very stable characteristics of an analytical method, because the blank signal and the signal generated by the very low concentrations of the analyte are frequently dependent on certain analytical parameters, including the purity of reagents, sample matrices, environmental conditions, instrumentation, and the analysts themselves. Sensitivity is a measure of the ability of an analytical method to discriminate between small differences in analyte concentration. It is defined as the analyte signal per unit concentration of the analyte. Despite the apparent simplicity of the sensitivity concept, a degree of confusion surrounds its use. This confusion stems from the perception that the sensitivity of a method is the same as the limit of detection. 

Waters of low-ionic strength (generally cation or anion totals <1 mg/litre) in which determinands may be close to or less than the limit of quantification. 

Polyfunctional mercaptans. Polyfunctional mercaptans are some of the most powerful odorants in nature, and they must be quantified at extremely low levels. The methods developed for their determination make use either of a selective separation using -hydroxy-mercurybenzoate (Tominaga et al. 1998 Tominaga and Dubourdieu 2006 Ferreira et al. 2007), of covalent chromatography (Schneider et al. 2003) or of the derivatization with pentaflurobenzyl bromide (Mateo-Vivaracho et al. 2006, 2007). A recent report using non-selective headspace isolation with SPME has also been published, but the limits of quantification are more than one order of magnitude above the odor thresholds (Fedrizzi et al. 2007). 

In a pilot study in six patients receiving intravenous flucytosine, hematotoxicity was monitored by measuring platelet and leukocyte counts flucytosine and 5-fluorouracil serum concentrations were measured using HPLC (19). The concentrations of 5-fluorouracil in the 34 available serum samples were below the limit of quantification (0.05 pg/ml), but flucytosine was detectable in all samples and the 5-fluorouracil metabolite, a-fluoro-P-alanine (FBAL), was detected at low concentrations in several samples. One patient developed thrombocytopenia (50 X 10 /1) during therapy, and one developed leukopenia (2.6 X 10 /1). The fact that 5-fluorouracil was not detected... 

The concentrations of topiramate have been measured in plasma and breast milk in five women with epilepsy during pregnancy and lactation (52). The umbilical cord plasma/maternal plasma ratios were close to unity, suggesting extensive transplacental transfer of topiramate. The mean milk/maternal plasma concentration ratio was 0.86 (range 0.67-1.1) at 2-3 weeks after delivery. The milk/maternal plasma concentration ratios at sampling 1 and 3 months after delivery were similar. Two of the breast-fed infants had detectable (>0.9 pmol/l) concentrations of topiramate 2-3 weeks after delivery, although they were below the limit of quantification (2.8 pmol/l), and one had an undetectable concentration. Thus, breastfed infants had very low topiramate concentrations. No adverse effects were observed in the infants. 

Repeatability is obtained when the analysis is performed in one laboratory by one analyst using the same equipment at the same day. Repeatability should be tested by the analysis of a minimum of five determinations at three different concentrations (low, medium and high) in the range of expected concentrations, according to FDA [16], However according to the ICH [79] repeatability could be measured by the analysis of three determinations at three different concentrations or through six determinations at 100 % of the test concentration. The latter one is for analysis when the concentration is supposed to be constant for all samples, e.g., pharmaceutical products. The acceptance criteria for precision depends much on the type of analysis. For compound analysis in pharmaceutical quality control, precision should be better than 2 % [82], For bioanalytical applications the precision values at each concentration level should be better than 15 % except for the lower limit of quantification (LLOQ) where is should not exceed 20 % [16], The intermediate precision shows the variations affected in day-to-day analysis, by different analysts, different instruments etc. Reproducibility, as above, represents the precision obtained between different laboratories. 

All metal concentrations found were relatively low and close to the limit of quantification of the SPEs/PalmSens method. As an example, cadmium concentrations estimated by SPEs/PalmSens and ICP/MS are presented in Figure 4.2.3. Although concentrations were over estimated they were in the same order of magnitude as the ICP/MS data. Results obtained by the SPEs/PalmSens are thus semi-quantitative but can lead to a comparison of a sample from one to another. Indeed, this screening resulted in the discrimination of the Oker River from the Aller River leading to the spatial variability assessment of the system. A representative picture of... 

---