Abstract
Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is a widely used quantitative method in small-molecule pharmacokinetic sample analysis. The linear dynamic range of mass analyzers, typically spanning 3 orders of magnitude, is usually insufficient for this purpose. Utilization of multiple isotopologues has been proposed as a compelling approach to expand the linear dynamic range of LC-MS/MS assays, particularly when the detector is saturated. Isotopologues are a statistical mixture of molecules of the same compound but of different exact masses due to the presence of natural chemical isotopes. While the concept of isotopologues is widely recognized in large-molecule bioanalysis and small-molecule metabolite profiling, it has not been commonly implemented in small-molecule targeted quantification. To increase the awareness of the value of isotopologues in small-molecule LC-MS/MS analysis, this minireview provides the basis of isotopologue distribution in MS/MS and summarizes published studies as well as our own experience in utilizing multiple isotopologues to expand the linear dynamic range of small-molecule LC-MS/MS assays. Considering that utilizing natural isotopologue transitions in the LC-MS/MS assays represents an easy, straightforward, and robust way to expand the linear dynamic range, we believe this method deserves wide application in small-molecule pharmacokinetic sample analysis and can particularly benefit people working in pharmacokinetic labs as well as the GLP bioanalytical labs in pharmaceutical industry.
Keywords: LCMS, isotopologues, natural isotopologue transitions, linear dynamic range, small-molecule pharmacokinetics
1. INTRODUCTION
Combining the separation power of liquid chromatography as well as the high specificity and sensitivity of mass spectrometer, liquid chromatography-tandem mass spectrometry (LC-MS/MS) has become the analytical method of choice in various applications, especially in small-molecule pharmacokinetics arena. The linear dynamic range of LC-MS/MS assays, a span from the lower limit of quantification (LLOQ) to the upper limit of quantification (ULOQ) of the calibration standards, usually covers 3 orders of magnitude for pharmacokinetic sample analysis. The importance of extending LLOQ is self-evident - without a sufficiently low LLOQ, drug concentrations may not be measurable at the later time points and consequently the terminal phase could be missed, leading to inaccurate estimate of drug half-life. In contrast, the importance of extending ULOQ is not readily understandable at the first glance since samples with concentrations higher than ULOQ can be diluted and reanalyzed, which does not seem to be a difficult task. As a result, during LC-MS/MS assay development phase, usually much more efforts are made to lower the LLOQ than to extend the ULOQ. When the ULOQ of a LC-MS/MS assay is relatively low, a substantial portion of the pharmacokinetic samples collected may need to be reanalyzed, which is both time- and labor-intensive. For those clinical pharmacokinetic labs which run large number of samples daily, the burden of sample reanalysis accumulates rapidly. One typical example is β-lactam antibiotics, such as cefazolin and piperacillin, which usually have concentrations >50 μg/mL in more than half of the collected samples following the standard dose regimens1. On the other hand, many published LC-MS/MS assays for these β-lactam antibiotics have low ULOQ (≤50 μg/mL)1,2, which means ≥50% of the samples need to be reanalyzed; this is unrealistic for those clinical pharmacokinetic labs. We encountered such situation and had to develop our own LC-MS/MS assays with desired ULOQ3 because the published methods did not meet our need. Another situation where sufficiently high ULOQ is important is in infant or animal pharmacokinetic studies, where sample volume is small and consequently sample reanalysis may not be feasible. Since both LLOQ and ULOQ are important, expanding the linear dynamic range is critical in order to detect low concentrations in samples collected at the terminal phase and meanwhile reliably measure high concentrations in samples collected at the initial phase without the need of sample reanalysis. In this minireview we review the strategy of using natural isotopologue signals in extending ULOQ and accordingly expanding the linear dynamic range of the LC-MS/MS assays for small-molecule sample analysis. Please note that this strategy does not improve LLOQ. The distribution of isotope peaks is a common knowledge in the field of LC-MS based proteomic research due to its use in deducing the charge state of a peptide peak (when resolution is sufficient) and thus the peptide mass. For small molecule bioanalysis, isotopic peak distribution is implemented in drug metabolite profiling, forensics, or doping control where high resolution mass spectrometer is utilized for molecule identification. In contrast, in the area of small molecule clinical or preclinical pharmacokinetics, where the focus of bioanalysis is targeted quantification, isotopic peak distribution has not been widely implemented by researchers. The aim of this mini-review is to increase the awareness of utilizing natural isotopologue transitions in LC-MS/MS assays, which represents an easy, straightforward and robust way to expand the linear dynamic range. We believe this method deserves wide application in small-molecule pharmacokinetic sample analysis and can particularly benefit people working in pharmacokinetic labs as well as the GLP bioanalytical labs in pharmaceutical industry. In this mini-review we first provide the background of isotopic distribution pattern as well as calculation of the probability of occurrence of isotopic peaks, then we summarize the key literature reports, together with our own experience, in utilizing natural isotopologue signals in expanding the linear dynamic range of the LC-MS/MS assays for small-molecule pharmacokinetic sample analysis.
2. PRINCIPLES OF ISOTOPE PEAK DISTRIBUTION IN MASS SPECTROMETRY
The principle of using isotopologue transitions to expand linear dynamic range is based on the distribution of naturally occurring chemical isotopes. For example, 98.9% of naturally found carbon, including those in our test compounds, is made up of the 12C isotope, and 1.1% naturally occurring carbon is 13C isotope. In addition to carbon, small-molecule compounds usually also contain oxygen, nitrogen, hydrogen, etc., all of which have their own isotope distribution as shown in Figure 1. This means that when we run samples in Q1 scan mode, we will get a statistical mixture of signals for our target compound – in addition to the highest peak which contains the most abundant isotopes, we will see a tail of peaks 1 Thomson (Th) apart (since small-molecules are usually singly charged). We have used this principle for extending the ULOQ of ampicillin and would like to use this compound as an example. Ampicillin has a formula of C16H19N3O4S with monoisotopic mass of 349.2 Da. Based on the isotope distribution listed in the left panel of Figure 1, we manually calculated the probability of ampicillin monoisotopic ion, [M-H]− (m/z 348.2 Th) and ampicillin isotopologue 1 ion, [+1M-H]− (m/z 349.2 Th), under negative ionization mode. Based on our calculations shown in the right panel of Figure 1, the probability of ampicillin [M-H]− and [+1M-H]− are 78.1% and 14.9%, respectively. From a different perspective, in any given population of ampicillin ions, 78.1% of the ions will have monoisotopic mass and 14.9% population will have isotopologue 1 mass. We did not perform manual calculation of isotopologue 2 or higher since the calculation is more complicated. In addition to manual calculation, prediction of isotopologue abundance can be performed by many web-based programs (e.g., https://www.sisweb.com/mstools/isotope.htm). It should be noted that the demonstrated calculation was for the parent ion only. Because the fragment ion also has its own distribution, the final signal for the +1parent > +1fragment channel will be reduced even further. Based on calculation of probability, for small molecules, the heavier the isotopologue, the lower its natural abundance. Therefore, monitoring the more abundant isotopologues will maintain sensitivity while monitoring the less abundant isotopologues will expand the ULOQ (since the less abundant isotopologues will not saturate the detector), and using multiple isotopologues together will allow expansion of the dynamic range.
Figure 1.
[Left panel] Nominal mass and abundance (%) of natural chemical isotopes of C, H, N, O and S. [Right panel] Calculations to predict the relative abundance of the precursor ions of ampicillin quantifier (m/z 348) and ampicillin isotopologue 1 (m/z 349).
3. KEY LITERATURE REPORTs OF THE UTILIZATION OF ISOTOPOLOGUES TO EXPAND THE LINEAR DYNAMIC RANGE OF SMALL MOLECULE LC-MS/MS ASSAYS
The earliest report of the use of isotopologues in expanding LC-MS/MS linear dynamic range was probably that by Liu et al., in 20114. In this study, an MRM assay was developed for the quantification of a flavor compound X (monoisotopic mass 343.2 Da) in rat plasma collected from a single dose pharmacokinetic study of X at 100 mg/kg. The assay was developed on an ABSciex QTRAP operating in positive ionization mode. Compound X was regularly quantitated using the most abundant transition of 344.2 → 136.1, which gave a linear response from 3 to 6,000 ng/mL. However, prior experience with an analog of X suggested that plasma concentration of X would likely be much higher than this ULOQ at earlier time points. Thus, the performance of the isotopologue transition of 345.2 → 137.1 was evaluated. Because the isotopologue is less abundant than the regular transition, the linear range was shifted to 150 – 60,000 ng/mL. By combining the regular transition and the isotopologue transition, the assay was able to cover a linear range of 3 – 60,000 ng/mL, a 10-fold increase in ULOQ compared to the assay with the regular transition alone. Indeed, the combination of the regular transition and the isotopologue transition in this case saved the researchers a lot of effort in sample reanalysis.
Following the successful application of multiple isotopologue transitions in expanding the linear dynamic range of LC-MS/MS assay on a QTRAP, Liu et al. later demonstrated the use of multiple isotopologues for quantifying other small-molecule drugs including diazinon (C12H21N2O3PS), imazapyr (C13H15N3O3), molinate (C9H17NOS), and thiabendazole (C10H7N3S) on a quadrupole time-of-flight (QTOF) system5. For each compound, three precursor ions were monitored: the monoisotopic ion ([M+H]+), isotopologue 1 ion ([+1M+H]+), and isotopologue 2 ion ([+2M+H]+ or [+3M+H]+). It is interesting that the relative abundance between the monoisotopic ion and its corresponding isotopologue ions were very similar among these four compounds. Specifically, with 1 Th increase in m/z, the abundance of isotopologue ions were 11 – 14%, and with 3 Th increase in m/z, the abundance of isotopologue ions were about 0.5% of the abundance of their respective monoisotopic ions. This is because most of the isotopic peaks are attributed to carbon isotopes, and the number of carbon atoms was similar among small molecule compounds. Corresponding to the difference in abundance, the ULOQ of isotopologue 1 and 2 ions were higher than the linear dynamic range of the monoisotopic ion, and using three ions together significantly expanded the linear dynamic range of the assay by 25 – 50 folds for each compound. The expansion of the linear dynamic range of diazinon based on multiple isotopologues are provided in Figure 2 as an example. The idea of expanding the linear dynamic range of LC-MS/MS assays using isotopologue transitions were later adopted by several other researchers, for example in the quantification of methamphetamine in urine6, in the assessment of 80 forensically relevant compounds in 11 different postmortem sample matrices7, and in the quantification of tivozanib in human and mouse plasma and mouse tissue homogenates8. A summary of literature reported applications of multiple isotopologue transitions in expanding LC-MS/MS assay linear dynamic range is provided in Table 1.
Figure 2.
The calibration curve of diazinon based on different isotopologues. The response of the normal (monoisotopic) ion is linear from 0.02 to 5 ng/mL. The response of the less sensitive isotopologues were linear from 0.1 to 20 ng/mL with m/z 306.112 and from 1 to 200 ng/mL with m/z 308.108. By monitoring multiple isotopologues, the normal linear dynamic range of diazinon was expanded from 0.02 – 5 ng/mL to 0.02 – 200 ng/mL. (Figure was adapted from Liu et al., 2014)
Table 1.
Literature report and our own data of the application of multiple isotopologues to expand the linear dynamic range (LDR) of LC-MS/MS assays for small molecule drugs.
| Reference | Instrument | Compound | Form of ion | m/z transition* | LDR (ng/mL) | LDR expansion | PMID |
|---|---|---|---|---|---|---|---|
| [+1M+H]+ | 345.2 → 137.1 | 150 – 60,000 | |||||
| [+2M+H]+ | 308.108 | 1 – 200 | |||||
| [+2M+H]+ | 264.126 | 2 – 500 | |||||
| [+3M+H]+ | 191.110 | 500 – 10,000 | |||||
| [+3M+H]+ | 205.043 | 5 – 200 | |||||
| [+1M+H]+ | 265.1 → 191.9 | 60 – 14,400 | |||||
| [+1M+H]+ | 300.1 → 200.1 | 50 – 2,000 | |||||
| [+1M+H]+ | 311.0 → 149.2 | 1000 – 24,000 | |||||
| [+1M+H]+ | 377.1 → 124.1 | 30 – 720 | |||||
| [+2M+H]+ | 138.1 → 121.1 | 5000 – 100,000 | |||||
| [+2M+H]+ | 152.1 → 121.0 | 5000 – 200,000 | |||||
| Bruin et al. | QQQ | [M+H]+ | 455 → 341 | 0.5 – 80 | 62.5 folds | 31352204 | |
| [+1M+H]+ | 456 → 341 | 50 – 5,000 | |||||
| Our study | QQQ | Oxfendazole | [M+H]+ | 316.2 → 191.4 | 0.5 – 2,500 | 8 folds | |
| [+1M+H]+ | 317.2 → 192.4 | 5 – 10,000 | |||||
| [+2M+H]+ | 318.2 → 193.4 | 50 – 20,000 |
The instrument was operated in data dependent acquisition. Quantification was based on the total intensity of all fragments generated by each precursor ion.
The assay was developed for 80 compounds. Due to space limit, only few compounds are included in this table.
4. OUR OWN EXPERIENCE WITH UTILIZING MULTIPLE ISOTOPOLOGUE IN THE LC-MS/MS ASSAYS OF OXFENDAZOLE AND ITS METABOLITES
To better understand the disposition of oxfendazole in mouse, we conducted a preclinical pharmacokinetics study in which either a single intravenous dose of 5 mg/kg or an oral dose of 15 mg/kg oxfendazole was administered in mice. Accordingly, a LC-MS/MS assay was developed for the simultaneous quantification of oxfendazole and its phase I metabolites, fenbendazole and oxfendazole sulfone, in mouse plasma based on our previously published assay for the quantification of oxfendazole in human plasma9. Detailed information including material, sample extraction and LC-MS/MS methods can be found from our report9. Originally, oxfendazole, fenbendazole and oxfendazole sulfone were subjected to positive ionization and quantified based on the most abundant monoisotopic (i.e., quantifier) transitions of m/z 316.2 → 191.4, 300.5 → 268.4, and 332.5 → 300.2, respectively. The presence of each analyte in mouse plasma was confirmed by monitoring the qualifier transitions of m/z 316.2 → 284.3, 300.5 → 159.2, and 332.5 → 158.9, respectively. Under these settings, the assay was linear from 0.5 to 2,500 ng/mL for oxfendazole and fenbendazole, and from 1 to 10,000 ng/mL for oxfendazole sulfone. The assay was able to capture all fenbendazole and oxfendazole sulfone concentrations. However, for oxfendazole, 50% samples collected after the intravenous dose and more than 75% samples collected after the oral dose had concentrations above the ULOQ. Because the original assay included qualifier transitions, which were less abundant than the quantifier transitions, we attempted to use oxfendazole qualifier to mitigate detector saturation and increase the ULOQ as suggested in several published studies10,11. However, this strategy failed. Even though oxfendazole qualifier was less abundant than oxfendazole quantifier, the difference was little. As a result, the ULOQ achieved with oxfendazole qualifier was still 2,500 ng/mL. Consequently, we resorted to the use of isotopologues. Two isotopologue transitions were added to the assay by adding 1 and 2 Th to both the precursor and product ions of oxfendazole quantifier. The linear dynamic range for oxfendazole based on isotopologue 1 (m/z 317.2 → 192.4) was 5 – 10,000 ng/mL and that based on isotopologue 2 (m/z 318.2 → 193.4) was 50 – 20000 ng/mL, which allowed quantification of oxfendazole in all samples without sample dilution.
Table 2 summarizes the expansion of oxfendazole linear dynamic range using multiple isotopologues. Overall, the calibration curve for oxfendazole consisted of 12 standards from 0.5 to 20,000 ng/mL. The assay was validated using 5 quality control (QC) levels of 1.5, 15, 750, 7,500 and 15,000 ng/mL so that there were 3 QC levels (low, medium and high) for the linear range of each isotopologue. QC average accuracies were in the range of 85.7 – 115.3% and coefficient of variations were less than 10%. QC falling in the overlapping regions of multiple calibration curves had different accuracies depending on which calibration curve was used. However, these differences were less than 15%. As demonstrated with oxfendazole LC-MS/MS assay, incorporation of multiple isotopologue is a simple and efficient way to resolve the issue of detector saturation and thus expand the linear dynamic range. Few additional calibrators were needed and only 1 additional QC level was required along with the addition of each isotopologue, because the medium and high QCs of the more abundant isotopologue can be the low and medium QCs of the less abundant isotopologue.
Table 2.
Calibration standards and quality control samples (N = 5 at each level) used to evaluate the performance of oxfendazole monoisotopic ion [M+H]+ and isotopologue ions [+1M+H]+ and [+2M+H]+.
| Concentration (ng/mL) | Calibration standards (% accuracy) | ||
|---|---|---|---|
| [M+H]+ | [+1M+H]+ | [+2M+H]+ | |
| 0.5 | 105 | ||
| 1 | 89.6 | ||
| 5 | 102 | 96.9 | |
| 10 | 106 | 104 | |
| 50 | 109 | 110 | 99.6 |
| 100 | 102 | 102 | 99.2 |
| 500 | 101 | 105 | 107 |
| 1000 | 89.4 | 99.6 | 102 |
| 2500 | 97.3 | 120* | 126* |
| 5000 | 73.5 | 94.1 | 103 |
| 10000 | 60.0 | 87.7 | 98.9 |
| 20000 | 41.9 | 77.4 | 89.8 |
| Quality controls (% accuracy)[%CV] | |||
| [M+H]+ | [+1M+H]+ | [+2M+H]+ | |
| 1.5 | 115.3 [6.8] | ||
| 15 | 108.1 [8.0] | 107.5 [4.9] | |
| 750 | 95.4 [2.7] | 105.0 [3.3] | 109.6 [2.8] |
| 7500 | 85.7 [3.5] | 96.9 [3.0] | |
| 15000 | 93.3 [2.5] | ||
Acceptance criteria for linearity: 75% of the calibration standards, including the LLOQ and ULOQ, have accuracy ranging from 85 to 115% (accuracy must be within 80–120% for LLOQ). This criteria is in line with the FDA guidance.
Shaded calibration standards either had very low S/N (i.e., not meeting the criteria for LLOQ) or had accuracy below 85% (i.e., cut of value for mass analyzer saturation).
samples were not included in linear regression because the accuracy was out of the 85 – 115% range
5. UTILIZATION OF MULTIPLE ISOTOPOLOGUES - ADDITIONAL CONSIDERATIONS
With the advancement of ionization technology, more ions are presented to the detector. Meanwhile, less improvement has been seen with mass detector. The increase in ionization together with the lack of improvement in detector means that saturation of detector would become more common with latter LC-MS/MS instruments. Consequently, undesired ULOQ and narrow dynamic range will likely continue to be issues frequently seen in small-molecule LC-MS/MS assays; this further highlights the value of utilizing isotopologue transitions to extend ULOQ and expand linear dynamic range. Isotopologues’ abundances and, correspondingly, their MS responses can be predicted based on the abundance of natural isotopes of primarily C and other atoms such as O and N, etc. If a priori knowledge of the drug concentration range in clinical samples are available, isotopologues can be easily included in the MS settings to expand the calibration range. Not much extra effort is required using this strategy as only 2–3 additional calibrators and 1 QC level are needed with the addition of each isotopologue.
Nonetheless, the inclusion of isotopologue transitions should be performed with caution. First, isotopologues can be used to address only saturation of detection but not saturation of ionization because each isotopologue has its own ion trajectory and thus reached the detector at distinguished time, but all isotopologues are ionized at the same time. Principles of ionization processes as well as strategies to avoid ionization saturation are beyond the scope of this mini-review. Information on the topic of source saturation can be found in our previous publication12. Another potential concern of using multiple isotopologues is when the compound used to make the calibration standards and the compound in the samples are from different sources, for example in the case of therapeutic drug monitoring (TDM). It was reported that active pharmaceutical ingredients synthesized by different pathways might have different isotopic composition13–15 due to either different isotopic composition of the raw materials or different chemical reactions favoring one isotope over the others (i.e., isotope fractionation16). The difference in isotopic compositions of compounds from different sources theoretically can lead to different distributions of isotopologues and inaccurate quantitative results. However, a study by Trobbiani et al.17 evaluating on variation of isotopic compositions of carbon, oxygen, hydrogen, nitrogen atoms showed that the potential for this quantitative error is fairly small (< 4%).Another potential concern is that including these additional isotopologue transitions may affect the quantification, particularly for an MRM assay with many analytes to begin with. In MRM mode, during the dwell time of 1 transition, all other transitions are lost in transmission. Therefore, inclusion of more transitions will lead to either of the two scenarios: 1) if dwell time is kept the same, the number of scans for each transition will be less, which leads to reduced data point and, consequently, less reliable peak integration; 2) if dwell time is reduced so that the number of scans for each transition is maintained, less ions will be collected for each transition, resulting in deterioration of the signal-to-noise ratio (S/N) and consequently the LLOQ. However, these potential caveats can easily be addressed by newer MRM algorithms such as Scheduled MRM (Analyst 1.6.1, AB Sciex) and Dynamic MRM (MassHunter Agilent) which can remove some transitions on the go based on the expected elution time of the compounds of interest, thus mitigate the effect of multiple transitions on S/N. An additional concern is the increased chance of carryover effect since much higher concentrations will be employed in the calibration standards, which potentially may lead to contamination especially for the samples analyzed after the highest concentration of calibration standard sample. If carryover effect does occur, a common strategy to address this issue is to run double black samples, once or even multiple times depending on the magnitude of the compound carryover on LC column, after the ULOQ calibration standard.
Another note is on how to report quantitative results. Because the calibration ranges of isotopologues overlap, samples with concentrations within this overlapping region might have multiple values depending on which calibration curve is used for quantification. Evaluation of samples in our oxfendazole preclinical study showed that the difference in concentrations of the same samples estimated using different isotopologue ranged from 0 to 20% for most samples (67/70 samples). In addition, these differences were neither concentration dependent nor isotopologue dependent. Thus, our recommendation is to either use the concentration estimated by the most abundant isotopologue applicable or use the average of estimated concentrations.
Examples in Section 3 (published reports) and Section 4 (our in-house data) clearly showed that utilizing natural isotopologue transitions in the LC-MS/MS assays represents an easy, straightforward, and robust way to expand the linear dynamic range. Although there are a number of potential concerns, most of them can be easily addressed. To the best of our knowledge, using natural isotopologue transitions in expanding the linear dynamic range of LC-MS/MS assay has not been applied in regulated bioanalytical work. We hope our manuscript could serve as a useful resource to facilitate the applications of this technique, especially by the GLP labs in pharmaceutical industry as time- and labor-intensive sample dilution and reanalysis could be avoided, and tremendous time, effort, and lab materials (e.g. blank plasma used for dilution) could be saved.
6. CONCLUSION
Utilization of multiple isotopologues is a straightforward yet very valuable tool to expand the linear dynamic range. Because the relative abundance of isotopologues is predictable, given a priori information of the range of drug concentration in biological samples, different isotopologues can be added to the MS/MS method without changing LC-MS/MS parameters. In terms of sample preparation, little extra effort is required as only a few extra calibrators and 1 QC level need to be added accompanying each isotopologue. On the other hand, linear dynamic range expansion using multiple isotopologue can save tremendous time and effort for sample dilution and re-analysis. With the technical improvement of ionization source and lack of advancement in detector, detector saturation is expected to happen more frequently. Therefore, consideration of multiple isotopologues during small-molecule LC-MS/MS assay development is strongly recommended.
ACKNOWLEDGEMENT
We would like to thank Dr. Jun Zhang for the critical review of our manuscript and valuable feedback. Partial support was provided by the Division of Microbiology and Infectious Disease, the National Institute of Allergy and Infectious Disease, the National Institutes of Health (HHSN272200800008C)
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Declaration of interests:
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
REFERENCES
- 1.Sime FB, Roberts MS, Roberts JA, Robertson TA 2014. Simultaneous determination of seven beta-lactam antibiotics in human plasma for therapeutic drug monitoring and pharmacokinetic studies. J Chromatogr B Analyt Technol Biomed Life Sci 960:134–144. [DOI] [PubMed] [Google Scholar]
- 2.Ohmori T, Suzuki A, Niwa T, Ushikoshi H, Shirai K, Yoshida S, Ogura S, Itoh Y 2011. Simultaneous determination of eight beta-lactam antibiotics in human serum by liquid chromatography-tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 879(15–16):1038–1042. [DOI] [PubMed] [Google Scholar]
- 3.D’Cunha R, Bach T, Young BA, Li P, Nalbant D, Zhang J, Winokur P, An G 2018. Quantification of cefepime, meropenem, piperacillin, and tazobactam in human plasma using a sensitive and robust liquid chromatography-tandem mass spectrometry method, part 1: assay development and validation. Antimicrob Agents Chemother 62(9). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Liu H, Lam L, Dasgupta PK 2011. Expanding the linear dynamic range for multiple reaction monitoring in quantitative liquid chromatography-tandem mass spectrometry utilizing natural isotopologue transitions. Talanta 87:307–310. [DOI] [PubMed] [Google Scholar]
- 5.Liu H, Lam L, Yan L, Chi B, Dasgupta PK 2014. Expanding the linear dynamic range for quantitative liquid chromatography-high resolution mass spectrometry utilizing natural isotopologue signals. Anal Chim Acta 850:65–70. [DOI] [PubMed] [Google Scholar]
- 6.Miller AM, Goggin MM, Nguyen A, Gozum SD, Janis GC 2017. Profiting from probability; combining low and high probability isotopes as a tool extending the dynamic range of an assay measuring amphetamine and methamphetamine in urine. J Anal Toxicol 41(5):355–359. [DOI] [PubMed] [Google Scholar]
- 7.Staeheli SN, Poetzsch M, Kraemer T, Steuer AE 2015. Development and validation of a dynamic range-extended LC-MS/MS multi-analyte method for 11 different postmortem matrices for redistribution studies applying solvent calibration and additional (13)C isotope monitoring. Anal Bioanal Chem 407(29):8681–8712. [DOI] [PubMed] [Google Scholar]
- 8.Bruin MAC, Rosing H, Lucas L, Wang J, Huitema ADR, Schinkel AH, Beijnen JH 2019. Development and validation of an LC-MS/MS method with a broad linear dynamic range for the quantification of tivozanib in human and mouse plasma, mouse tissue homogenates, and culture medium. Journal of Chromatography B 1125. [DOI] [PubMed] [Google Scholar]
- 9.Bach T, Bae S, D’Cunha R, Winokur P, An G 2019. Development and validation of a simple, fast, and sensitive LC/MS/MS method for the quantification of oxfendazole in human plasma and its application to clinical pharmacokinetic study. J Pharm Biomed Anal 171:111–117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Curtis MA, Matassa LC, Demers R, Fegan K 2001. Expanding the linear dynamic range in quantitative high performance liquid chromatography/tandem mass spectrometry by the use of multiple product ions. Rapid Commun Mass Spectrom 15(12):963–968. [DOI] [PubMed] [Google Scholar]
- 11.Yuan L, Zhang D, Jemal M, Aubry AF 2012. Systematic evaluation of the root cause of non-linearity in liquid chromatography/tandem mass spectrometry bioanalytical assays and strategy to predict and extend the linear standard curve range. Rapid Commun Mass Spectrom 26(12):1465–1474. [DOI] [PubMed] [Google Scholar]
- 12.An G, Bach T, Abdallah I, Nalbant D 2020. Aspects of matrix and analyte effects in clinical pharmacokinetic sample analyses using LC-ESI/MS/MS - Two case examples. J Pharm Biomed Anal 183:113135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Jasper JP, Westenberger BJ, Spencer JA, Buhse LF, Nasr M 2004. Stable isotopic characterization of active pharmaceutical ingredients. J Pharm Biomed Anal 35(1):21–30. [DOI] [PubMed] [Google Scholar]
- 14.Jasper JP, Weaner LE, Duffy BJ 2005. A preliminary multi-stable-isotopic evaluation of three synthetic pathways of Topiramate. J Pharm Biomed Anal 39(1–2):66–75. [DOI] [PubMed] [Google Scholar]
- 15.Wokovich AM, Spencer JA, Westenberger BJ, Buhse LF, Jasper JP 2005. Stable isotopic composition of the active pharmaceutical ingredient (API) naproxen. J Pharm Biomed Anal 38(4):781–784. [DOI] [PubMed] [Google Scholar]
- 16.Sharp Z 2017. Principles of stable isotope geochemistry. 2nd ed. [Google Scholar]
- 17.Trobbiani S, Stockham P, Scott T 2017. Increasing the linear dynamic range in LC-MS: is it valid to use a less abundant isotopologue? Drug Test Anal 9(10):1630–1636. [DOI] [PubMed] [Google Scholar]