Xi Zhang†
a,
Xinbei Jia†bc,
Weihang Tongd,
Hui Chena,
Ning Leid,
Guangrun Lia,
Jun Tai*b and
Pengfei Li*a
aPharmacy Department of Beijing Chaoyang Hospital, Capital Medical University, No. 8 Gongti South Road, Chaoyang District, Beijing, China. E-mail: lee-pf@163.com
bDepartment of Otorhinolaryngology, Children's Hospital, Capital Institute of Pediatrics, No. 2 Yabao Road, Chaoyang District, Beijing, China. E-mail: trenttj@163.com
cChinese Academy of Medical Sciences, Peking Union Medical College, No. 30 Shuangqing Road, Haidian District, Beijing, China
dPharmacy Department of PLA Rocket Force Characteristic Medical Center, No. 16 Xinwai Street, Xicheng District, Beijing, China
First published on 18th November 2022
Clofarabine is approved for the treatment of relapsed or refractory acute lymphoblastic leukemia (ALL) in pediatric patients aged 1 to 21 years. Its pharmacokinetic (PK) exposure is strongly related to clinical outcomes and high risk of adverse reactions. PK-guided dosing of nucleoside analogs has the potential to improve survival and reduce toxicity in children. Considering that blood collection is an invasive operation and that the volume of blood collected is usually limited in pediatric ALL patients, a convenient and efficient method for the quantification of clofarabine in human urine and plasma was established with an LC-MS/MS system. Standard curves were shown to be liner in the range of 2.00–1000.00 ng mL−1 in both urine and plasma. Analytical validation of the assay included the assessment of linearity, accuracy (RE: −6.62% to 2.32%), intra-assay precision (RSD: 0.81% to 3.87%) and inter-assay precision (RSD: 1.88% to 5.69%). The absolute recovery rates of clofarabine were 85.50 ± 4.80%, 89.40 ± 0.70% and 98.00 ± 0.40% in urine and were 80.76 ± 1.88%, 86.81 ± 0.75%, 88.10 ± 0.61% in plasma at 5.00, 30.00 and 800.00 ng mL−1, respectively. The selectivity, stability and matrix effects conformed to the biological sample analysis requirements. The cumulative urine excretion rates for 24 hours of the three children with relapsed and refractory acute lymphoblastic leukemia were 72.22%, 87.88%, 82.16%, respectively. The PK data of the pediatric patient numbered lflb13-05 are very inconsistent with that of the other two children subjects, demonstrating that there may be an individual variation in Chinese pediatric patients, so the dose should be individualized based on the monitoring of drug concentration. The method is convenient, sensitive, and accurate, and it is suitable for the determination of clofarabine urine and plasma concentration. This is the first report on the pharmacokinetics of clofarabine in Chinese ALL children. Furthermore, it could be an alternative method to clinical monitoring of clofarabine.
In previously reported studies,2–5 patient responses to the clofarabine-based treatment regimens were diverse, and there was a high risk of infection, hepatotoxicity, and death from treatment-related adverse effects. Toxic effects of clofarabine include liver function abnormalities, dose-limiting toxicities of hand-foot syndrome, skin toxicity, and systemic inflammatory response syndrome.6–8 The study by Bozena Büttner9 showed that with increased individual clofarabine exposure, the risk of damage to hepatocytes measured by elevated liver enzymes increased, but antileukemic efficacy did not increase. Clofarabine can also cause acute kidney injury (AKI), the established AKI profile for clofarabine occurred in 16% of patients.10 Increased exposure to clofarabine is associated with an increased risk of renal toxicity in older adults.11
Vijay Ivaturi12 reported that PK-guided dosing of nucleoside analogs has the potential to improve survival and reduce toxicity in children at high risk for graft rejection and disease relapse. Therefore, personalized assays such as PK evaluation of systemic clofarabine exposures are urgently important. Clofarabine population PK for ALL or HCT in pediatric patients from the Netherlands2 and America13 have been reported, however, there were no PK data from Chinese pediatric patients.
Plasma concentrations were usually determined for routine therapeutic drug monitoring (TDM). Current practice includes monitoring clofarabine elimination via serum concentration, a lot of approaches have been reported for the determination of clofarabine concentration in blood.1,13–16 However, considering that the blood collection is an invasive operation, this process could easily cause infection in the patient,17 and the volume of blood collected is usually limited in pediatric patients with ALL.18 The clinical application of these studies is limited. In recent years, the role of biomarkers (e.g., urine) is actively being investigated, drug exposures could be robustly back-calculated by the known urinary excretion fraction.19 The literature20 reported that the urine monitoring system would allow for immediate and accurate prediction and intervention to prevent methotrexate-induced AKI. Approximately 60% of clofarabine is excreted from urine, it is mostly cleared from renal system. Therefore, the level of clofarabine in the urine is physiologically correlated with the circulation of clofarabine in the blood. However, urine collection is noninvasive and can prevent infection during the sample collection and blood consumption in pediatric ALL patients, so the quantification of clofarabine in urine has much more benefits than the determination in other sample matrices. Perhaps, dynamic monitoring of urine clofarabine concentration in patients could be an alternative to clinical monitoring.
In this study, a sensitive and convenient LC-MS/MS method was developed for the first time for the determination of clofarabine in human urine, and the method for the quantification of clofarabine in plasma was established at the same time. The calculation curve, selectivity, accuracy, precision, stability, matrix effect, and recovery were fully validated. The proposed method was successfully applied to three pediatric Chinese ALL patients.
Gradient elution chromatography was carried out at 35 °C on a 150 mm × 4.6 mm, 5 μm Aglient TC-C18 column (Agilent Technologies, Palo Alto, CA, USA), and maintained at a flow rate of 1.0 mL min−1. The mobile phase consisted of methanol and 1 mM aqueous ammonium acetate. The proportion of methanol gradually increased from 10% to 95% during the first 0.90 min. During 0.90 to 2.49 min, the proportion of methanol remained 95%. After that, methanol represented 10% until 5.00 min. The column effluent was split and the speed of which entered the mass spectrometer was approximately 0.5 mL min−1. The injection volume was 10 μL.
The ESI source in positive-ion mode was selected for all experiments. The LC-MS/MS detector was operated at medium resolution in MRM mode using the mass transition ion pairs m/z 304.0 → 170.0 for clofarabine, m/z 515.1 → 276.1 for telmisartan. To optimize the MS parameters, we used a syringe pump for infusing a standard solution of analyte and IS into the mass spectrometer. The optimized parameters were as follows: curtain gas, gas 1 and gas 2 (nitrogen) 25, 45 and 55 units, respectively; dwelling time 200 ms; source temperature 550 °C; ion spray voltage 5500 V. Declustering potential (DP) and collision energy (CE) were 32 V and 29 eV for clofarabine, 35 V and 50 eV for telmisartan, respectively. The collision gas and interface heater were set to ‘medium’ and ‘on’ mode.
Calibration standards and QC samples (6 samples per day per concentration level) for urine and plasma were analyzed for at least three separate days. The linearity of the calibration curves which were based on peak areas were assessed by weighted (1/x2) least-squares linear integral analysis (y = ax + b as regression model). The lower limits of quantification (LLOQ) of clofarabine in urine or plasma were the lowest limit of the calibration curves and meet the requirement that the mass spectrum signal was greater than or equal to 10 signal-to-noise ratio. Relative standard deviation (RSD) represented intra- and inter-day precision, and relative error (RE) was used to assess accuracy. Absolute recovery rates of clofarabine were measured by comparing the peak areas of extracted QC samples with those of reference QC solutions reconstituted in blank urine or plasma. Matrix effect was evaluated by comparing peak areas of QC and IS solutions reconstituted in blank urine or plasma extracts with those of the same solutions injected directly into the LC-MS/MS system.
The stabilities of clofarabine both in urine and plasma were evaluated by the spiked QC samples with six replicates under the following conditions: one freeze–thaw cycle and three freeze–thaw cycles at −20 °C, long term at −20 °C for 20, 50, 90 days, short-term and autosampler stabilities at room temperature for 24 h. Samples for stability tests were quantified using freshly prepared calibration standards.
To investigate the suitability of this method for high concentration samples that exceeded the upper limit of detection, high concentration plasma samples were diluted to 6 replicates of dilution samples, and the RSD was evaluated. Each high concentration urine or plasma sample contained clofarabine (15 μg mL−1) and was diluted 500 times with blank urine or plasma.
The developed LC-MS/MS method was used in urine and plasma samples of three pediatric ALL patients. The pharmacokinetic parameters of clofarabine were calculated. The cumulative urinary excretion curves and the concentration–time curves of clofarabine were dawn.
Although instrumental tests with standard solutions in the positive ion detection mode produced a higher response than in the negative mode, presumably because of the high proton affinity of clofarabine. But the signal intensity was not the most ideal. 1 mM ammonium acetate16 or the mixture of 2 mM ammonium acetate and formic acid1 were reported to increase the intensity of the signal compared to the neutral mobile phase solvents. To facilitate fully evaporation of ionized analytes at the ion source, different concentrations of formic acid and ammonium acetate were tested, and 1 mM aqueous ammonium acetate was finally adopted as the mobile phase.
The clofarabine and telmisartan spectra contained predominant protonated molecules at m/z 304.0 and 515.1 in full-scan positive mode, respectively. The product ion spectra of [M + H]+ showed fragment ions of clofarabine at m/z 170.0 and those of telmisartan at m/z 276.1 (Fig. 1) were dominantly presented and finally chosen for the acquisition of multiple reaction monitoring (MRM) of clofarabine and telmisartan, respectively.
For the determination of the compound in biological samples by LC-MS/MS, sufficient chromatographic resolution is necessary to avoid possible interferences from drug-related biotransformation products.22 As many salts in urine are also polar compounds, achieving good resolution separation and high sensitivity of clofarabine is even more challenging; therefore, the composition of the mobile phase was optimized through several trials. Ammonium acetate is an organic salt; the buffer solution prepared from ammonium acetate was more stable than that from formic acid. At the same time, ammonium acetate can ensure better ionization efficiency of clofarabine, so a mobile phase consisted of methanol and 1 mM ammonium acetate in water was finally adopted. Gradient elution was final selected because it provides better chromatographic resolution and better loading capacity than isocratic elution. In order to avoid the disadvantage of longer analysis time for column re-equilibration of gradient elution, the concentration ratio of methanol was constantly changed from 10% to 95% during 0–0.9 min, and the flow rate was 1.0 mL min−1. Under optimized conditions, there were no co-elution interferences, and the retention time was typically 3.71 min for clofarabine, and 3.72 min for telmisartan. The cycle time was 5.00 min allowing a sample throughput of 200–250 samples per day. This newly established method has a substantially shorter total run time than the reported methods13,15 and a similar total run time with several methods,1,14,16 which are used only for the quantification of clofarabine in plasma. The present validated method is the first validated method for the quantification of clofarabine in urine.
Based on our previous experience, both telmisartan and ampicillin had good ionization ability in the positive ionization mode. Comprehensively considering the stability under the sample preparation and analysis procedure, telmisartan was finally chosen as IS after several trials.
The typical calibration curve equation for the urine sample was y = 0.00306x + 0.00103, and the correlation coefficient was 0.9989. The calibration curve concentration range was 2.00–1000.00 ng mL−1 with the LLOQ of 2.00 ng mL−1 for urine.
As Table 1 shows, the intra- and inter-day RSDs of the urine QC were 0.81–3.59%, 1.88–4.12%, the REs were −6.33–2.32%, −6.62–1.23%. All values were within the acceptable range.
Sample | Concentration (ng mL−1) | Intra-day | Inter-day | ||||
---|---|---|---|---|---|---|---|
Measured (ng mL−1, x ± s) | RSD (%) | RE (%) | Measured (ng mL−1, x ± s) | RSD (%) | RE (%) | ||
a Abbreviations: RSD: relative standard deviation; RE: relative error. | |||||||
Urine | 5.00 | 5.01 ± 0.18 | 3.59 | 0.28 | 5.06 ± 0.21 | 4.12 | 1.23 |
30.00 | 28.10 ± 0.43 | 1.53 | −6.33 | 28.01 ± 0.53 | 1.88 | −6.62 | |
800.00 | 818.60 ± 6.61 | 0.81 | 2.32 | 778.55 ± 31.02 | 3.98 | −2.68 | |
Plasma | 5.00 | 4.90 ± 0.19 | 3.87 | −1.94 | 4.85 ± 0.28 | 5.69 | −3.08 |
30.00 | 29.22 ± 0.80 | 2.75 | −2.60 | 29.01 ± 1.62 | 5.60 | −3.30 | |
800.00 | 809.03 ± 8.09 | 1.00 | 1.13 | 806.86 ± 36.48 | 4.52 | 0.86 |
The absolute recovery rates of clofarabine were 85.50 ± 4.80%, 89.40 ± 0.70% and 98.00 ± 0.40% in urine at 5.00, 30.00 and 800.00 ng mL−1, respectively. The results of the matrix effects indicated that no co-eluting endogenous substances significantly influenced the ionization of clofarabine and the internal standard.
The stability results (Table 2) suggested that the REs met the requirements, clofarabine in urine was stable under all conditions. The results of the stability of the dilution showed that the recovery rate of clofarabine was 96.73 ± 1.94% in urine. Thus, this determination method was not only suitable for samples within the detection limit, but also was superior for high-concentration samples that exceeded the upper limit of detection.
Sample | Concentration (ng mL−1) | 1 freeze–thaw | 3 freeze–thaw | Long term at −20 °C for 20 days | Long term at −20 °C for 50 days | Long term at −20 °C for 90 days | Short-term and autosampler at room temperature for 24 h | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Measured (ng mL−1) | RE (%) | Measured (ng mL−1) | RE (%) | Measured (ng mL−1) | RE (%) | Measured (ng mL−1) | RE (%) | Measured (ng mL−1) | RE (%) | Measured (ng mL−1) | RE (%) | |||
a Abbreviations: RE: relative error. | ||||||||||||||
Urine | 5.00 | 4.85 ± 0.12 | −3.07 | 5.25 ± 0.08 | 4.97 | 5.32 ± 0.14 | 2.69 | 5.44 ± 0.21 | 8.84 | 5.16 ± 0.20 | 3.14 | 5.01 ± 0.08 | 0.15 | |
30.00 | 32.17 ± 0.17 | 7.23 | 28.96 ± 0.64 | −3.45 | 27.33 ± 0.59 | 2.17 | 27.93 ± 1.40 | −6.90 | 33.49 ± 0.49 | 11.63 | 28.43 ± 1.18 | −5.23 | ||
800.00 | 715.15 ± 9.62 | −10.61 | 750.26 ± 12.26 | −6.22 | 765.31 ± 7.03 | 0.92 | 766.33 ± 7.58 | −4.21 | 885.68 ± 11.45 | 10.71 | 800.34 ± 21.34 | 0.04 | ||
Plasma | 5.00 | 5.09 ± 0.26 | 1.77 | 5.02 ± 0.09 | 0.34 | 4.92 ± 0.18 | −1.56 | 4.88 ± 0.10 | −2.49 | 4.81 ± 0.17 | −3.71 | 4.77 ± 0.15 | −4.51 | |
30.00 | 27.29 ± 0.37 | −9.05 | 29.64 ± 0.42 | −1.21 | 29.31 ± 0.57 | −2.30 | 32.53 ± 1.38 | 8.42 | 30.00 ± 1.90 | 0.02 | 27.97 ± 1.19 | −6.75 | ||
800.00 | 700.94 ± 12.64 | −12.38 | 841.04 ± 20.48 | 5.13 | 874.00 ± 15.15 | 9.25 | 815.57 ± 17.73 | 1.95 | 802.50 ± 17.71 | 0.31 | 815.95 ± 19.72 | 1.99 |
The typical calibration curve equation for the plasma sample was y = 0.00385x + 0.00195, and the correlation coefficient was 0.9949. The calibration curve concentration range was 2.00–1000.00 ng mL−1 with the LLOQ of 2.00 ng mL−1 for plasma. The previous studies reported that the LOD of clofarabine in plasma were 2.50 ng mL−1 (ref. 14 and 16) or 20.00 ng mL−1.1 Compared with the previous studies, the established method in plasma has a lower LOD. In addition, the detection range of Liusheng Huang's13 study was 0.50–80.00 ng mL−1, although the LOD is lower than that of this study, the detection range of this study is wider and more suitable for the needs of clinical sample testing.
As Table 1 shows, the intraday and inter-day RSDs of plasma QC were 1.00–3.87%, 4.52–5.69%, RE were −2.60–1.13%, −3.30–0.86%. Both the RSDs and the REs were less than 15%.
The absolute recovery rates of clofarabine were 80.76 ± 1.88%, 86.81 ± 0.75%, 88.10 ± 0.61% in plasma at 5.00, 30.00 and 800.00 ng mL−1, respectively. The results of the matrix effects indicated that no co-eluting endogenous substances significantly influenced the ionization of clofarabine and the internal standard.
The stability results (Table 2) suggested that the REs in plasma all met the requirements, plasma sample was stable under all conditions. The results of the stability of the dilution showed that the recovery rate of clofarabine was 101.50 ± 0.58% in plasma. Thus, this determination method can meet the needs of clinical detection of higher concentration samples.
Patient ID | lflb13-02 | lflb01-01 | lflb13-05 | |||
---|---|---|---|---|---|---|
PK parameters | Single-dose | Multiple-dose | Single-dose | Multiple-dose | Single-dose | Multiple-dose |
a Abbreviations: Tmax: time of Cmax; Cmax: peak plasma concentration; AUC: area under the curve; Vd: the apparent volume of distribution; T1/2: terminal half-life; MRT: mean residue time; CL: clearance. | ||||||
Tmax (h) | 2.50 | 2.50 | 2.50 | 2.50 | 2.50 | 2.50 |
Cmax (ng mL−1) | 565.35 | 921.55 | 800.28 | 1675.00 | 524.58 | 513.84 |
AUC0−t (h ng mL−1) | 3258.70 | 5193.25 | 5392.26 | 5156.59 | 2238.86 | 2503.96 |
AUC0−INF (h ng mL−1) | 3295.17 | 5314.66 | 5434.06 | 5237.99 | 2320.29 | 2570.19 |
Vd (L) | 0.10 | 0.09 | 0.07 | 0.11 | 0.19 | 0.18 |
T1/2 (h) | 3.64 | 5.14 | 3.16 | 5.07 | 5.67 | 5.82 |
MRT (h) | 5.59 | 100.11 | 5.98 | 99.73 | 4.99 | 100.23 |
CL (L h−1) | 0.02 | 0.01 | 0.01 | 0.02 | 0.02 | 0.02 |
Fig. 3 Cumulative urinary excretion curves (a) and plasma concentration–time curve (b) of clofarabine after clofarabine injection. |
Previous literature27 reported that renal function and body weight are two important covariates for clearance, therefore, they should be fully considered as components when determining the individual dose of clofarabine.27 For a basis for optimizing the schedule of clofarabine administration, urine concentration should be standardized by the basal creatinine level. At present, the quantitative method ultimately remains qualitative in that clofarabine can relate only to positive or negative aspects of treatment. Future research should focus on the relationship between serum and urine clofarabine concentrations and determining approximate therapeutic concentration ranges of urine to facilitate safety and efficacy.
Footnote |
† These authors contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2022 |