You are here

Article Text

Article menu

Download PDFPDF

Research
Original article
Phenotypical difference in deamination of cytarabine is not evident in induction therapy for acute myeloid leukaemia
  1. Mikkel Krogh-Madsen1,
  2. Steen Hansen Honoré2,
  3. Morten Krogh Jensen3,
  4. Ove Juul Nielsen4,
  5. Hannah Søndergaard Jensen1,
  6. Per Hartvig Honoré1
  1. 1Faculty of Health and Medical Sciences, Department of Drug design and Pharmacology, University of Copenhagen, Denmark
  2. 2Faculty of Health and Medical Sciences, Department of Pharmacy, University of Copenhagen, Denmark
  3. 3Department of Haematology L-124, Herlev Hospital, Herlev, Denmark
  4. 4Department of Haematology L-4042, Rigshospitalet, Copenhagen, Denmark
  1. Correspondence to Professor Per Hartvig Honoré, Faculty of Health and Medical Sciences, Department of Drug design and Pharmacology, University of Copenhagen, Universitetsparken 2, DK-2100 Copenhagen, Denmark; peh{at}sund.ku.dk

Abstract

Objective To investigate the uracil arabinoside/cytarabine (Ara-U/Ara-C) ratios with the lower dose in adult acute myeloid leukaemia (AML) induction therapy (100 mg/m2 Ara-C) where no enzyme saturation is expected.

Methods A precise and robust high-performance liquid chromatography (HPLC) method for simultaneous determination of Ara-C and its main inactive metabolite Ara-U in human plasma was developed and validated. Nineteen patients with acute myeloid leukaemia were treated with Ara-C in a dose of 100 mg/m2 together with daunorubicin and etoposide. Plasma concentrations were used to construct the standard normality plot to indicate towards two different phenotypes for the deamination enzyme. This was confirmed with the Shapiro–Wilks test for normality and a histogram of the distribution of the ratios.

Results The lower limits of quantification (LLoQ) of the developed method were 32 ng/ml and 10 ng/ml for Ara-C and Ara-U, respectively. Precision, accuracy, recovery, selectivity and stability varied by no more than 15% at concentrations above LLoQ and by 20% at LLoQ, except for long-term stability of Ara-U. Both the Shapiro–Wilks test for normality and the histogram showed a unimodal distribution. The non-transformed values of the Ara-U/Ara-C ratios were between 0.3 and 17.7 (median 2.2). No correlations between Ara-U/Ara-C ratios and age, sex, liver or renal function or treatment outcome were found. Fifteen of the 19 patients had complete remission of the disease and 2 had partial remission.

Conclusions Division into slow and fast Ara-C metabolisers in this patient population could not be made and specific dose individualisations can therefore not be recommended.

  • PHARMACOTHERAPY
  • SIDE EFFECTS OF DRUGS

https://doi.org/10.1136/ejhpharm-2012-000270

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Cytarabine (Ara-C) has been used in the treatment of acute myeloid leukaemia (AML) for decades.1 It is rapidly deaminated in the plasma by the enzyme deoxycytidine deaminase to the inactive metabolite uracil arabinoside (Ara-U). A previous study has outlined the existence of two phenotypes for the deamination process and suggested a link between phenotype and patient outcome.2 The study used the ratio of Ara-U/Ara-C peak plasma concentrations to distinguish between fast and slow metabolisers. A later study has not been able to confirm these results.3 The previous studies were both performed on high-dose Ara-C (1–3 g/m2). With high-dose Ara-C the plasma concentration reaches a level in some patients where the enzyme systems for deamination and deoxycytidine kinase may be saturated. The first step of the intracellular phosphorylation of Ara-C by deoxycytidine kinase is saturated at concentrations above 10 µM (ie, 2430 ng/ml).4 This concentration level is easily exceeded with high-dose Ara-C, but usually not by the doses in conventional dose therapy (100 mg/m2).5 This could influence the Ara-U/Ara-C ratio and thus disturb the image of slow versus fast metabolisers.

    A study to investigate the relationship between Ara-C and its main metabolite Ara-U was performed. The aim of the study was to investigate the ratios with the lower dose in adult AML induction therapy (100 mg/m2 Ara-C) where no enzyme saturation is expected.6

    Materials and methods

    Nineteen adult patients newly diagnosed with AML were included in the study (11 men, 8 women). The patients were diagnosed according to the WHO classification with >20% myeloblasts in the bone marrow. The patients were recruited from two hospitals in Denmark (Rigshospitalet and Herlev Hospital) and received 10 days of treatment with Ara-C (100 mg/m2 twice daily on days 1–10), 3 days of treatment with daunorubicin (50 mg/m2 daily on days 1, 3 and 5) and 5 days of treatment with etoposide (100 mg/m2 daily on days 1–5). Ara-C was given as an intravenous push over 5 min, while both daunorubicin and etoposide were given as 1 h infusions. All patients were given adequate transfusion therapy and were treated for the numerous complications that occur in this patient population in accordance with the normal procedures at the two hospitals.

    A method for the simultaneous quantification of Ara-C and Ara-U with high-performance liquid chromatography-ultraviolet (HPLC-UV) was developed and validated based on Food and Drug Administration (FDA) guidelines.7 The HPLC equipment was from the Agilent 1120 Compact LC series. It consisted of a UV detector, oven, autosampler, pump and degasser. The software used for analysis of the signals was E2Chrom Elite Compact (Agilent, Denmark). The column was an Acclaim Polar Advantage II (4.6×150 mm, 3 µm) C18 column (Dionex, Denmark). Ara-C was purchased from Sigma-Aldrich (Denmark) and Ara-U was purchased from Santa Cruz Biotechnology (USA). Tetrahydrouridine (THU) (VWR, Denmark) was added to the plasma samples to inhibit the deaminase activity.8 All other chemicals and solvents were of analytical grade.

    Ara-C and Ara-U were diluted in 5 mM ammonium formate (pH 4.6) to a concentration of 0.1 mg/ml and kept at −20°C. Calibration curves were made from further dilutions of these stock solutions. Spiked plasma samples were made with plasma from healthy volunteers. The volume of analyte stock solution in the spiked plasma did not exceed 10% of the total sample volume.

    Blood samples from the included patients with AML were collected in heparinised tubes containing THU (0.1 mg THU/ml blood). Plasma was isolated by centrifugation at 2.500×g for 15 min and kept at −20°C until analysis. Before HPLC analysis, 500 µl plasma was mixed with 500 µl of a solution containing 0.05% v/v heptafluorobutyric acid in acetonitrile and was kept on ice for 10 min in order to precipitate plasma protein. The samples were then centrifuged at 11.000×g for 5 min and the supernatant was evaporated to dryness at 40°C under a gentle stream of nitrogen. The residue was redissolved in 250 µl mobile phase A (0.05% v/v heptafluorobutyric acid in Milli-Q water). The samples were quantified with gradient elution reverse-phase chromatography using the mobile phases A and B (methanol). The flow rate was 1 ml/min, the column temperature was 25°C and the injected sample volume was 50 µl. The wavelength was changed in each run from 262 nm (optimal for Ara-U) to 280 nm (optimal for Ara-C). Calibration curves from spiked plasma samples were made on each day of patient plasma sample analysis.

    The Ara-U/Ara-C ratio was determined with the measured concentration of the blood samples which were drawn less than 15 min after the end of the infusion. The analysis of the distribution of Ara-U/Ara-C ratios to identify groups of patients showing different rates of deamination was performed using a histogram, a normal probability plot and the Shapiro–Wilks test for normality. The Shapiro–Wilks test tests the null hypothesis that a group of n samples come from a normally distributed population.9 These were applied with the statistical software R. Two-tailed statistics and significance levels of 0.05 were used throughout. The log of the ratios was used in the analysis since distributions of ratios are often skewed and a log transformation can result in a nearly normal distribution.2 Clearly, this nearly normal distribution will not be nearly normal if the distribution truly is a multimodal distribution, in which case a phenotypic distinction can be made. Single time point measurements were evaluated on the Ara-U/Ara-C ratio and linked to age, sex, liver and renal function and treatment outcome. Treatment outcome was defined as remission status approximately 1 month after induction treatment. Remission status was either complete remission (<5% blasts in normocellular bone marrow), partial remission (blast count reduced by at least half to a value between 5% and 15% in otherwise normocellular bone marrow) and resistant disease (the bone marrow showed persistent AML). Renal function was estimated with calculated creatinine clearances using the Cockcroft–Gault formula based on serum creatinine measurements. Alanine aminotransferase was used as a measure of liver damage and albumin and bilirubin as surrogates for liver functionality.

    Results

    The developed method was selective and had a lower limit of quantification (LLoQ) for Ara-C of 32 ng/ml and for Ara-U of 10 ng/ml (figure 1). Precision, accuracy, recovery and stability (freeze-thaw stability, stability in whole blood, post-preparative stability and long-term stability in plasma) varied by no more than 15% at concentrations above LLoQ and by 20% at LLoQ (table 1), except for long-term stability of Ara-U in patient plasma. Five different plasma samples from one patient were analysed with 2 months in between each analysis for long-term stability testing. The concentration relative to that at time zero was 92.9–111.6% (mean 98.5%) for Ara-C and 105.2–152.5% (mean 128.8%) for Ara-U. The Ara-C plasma concentrations measured with this method were compared with the concentrations of the same samples measured with a different method10 and were found to be similar (data not shown).

    View this table:
    Table 1

    Data from the validation procedure for the Ara-C and Ara-U HPLC method

    Figure 1
    Figure 1

    Chromatogram of a blank plasma sample (solid line) and a spiked plasma sample (500 ng/ml of both analytes, dotted line). Ara-U, uracil arabinoside; Ara-C, cytarabine.

    The standard normality plot indicated no tendency towards two different phenotypes for the deamination enzyme (figure 2). This was confirmed with the Shapiro–Wilks test for normality (p=0.58) and a histogram of the distribution of the ratios (figure 2). Both showed a unimodal distribution. The non-transformed values of the Ara-U/Ara-C ratios were between 0.3 and 17.7 with a median value of 2.2. These values were similar to the values obtained by Burk et al3 who found ratios between 1.74 and 13.9, and lower than the values in the study by Kreis et al where 29% of the patients had a ratio above 14.2

    Figure 2
    Figure 2

    Standard normality plot (left) and histogram (right) of uracil arabinoside/cytarabine (Ara-U/Ara-C) ratios. The log of the ratios was used to make the plots.

    The demographic characteristics of the patients are summarised in table 2. Fifteen of the 19 patients were considered to have achieved complete remission, 2 had partial remission and 2 had resistant disease. No correlations between Ara-U/Ara-C ratios and age, sex, liver or renal function or treatment outcome were found.

    View this table:
    Table 2

    Patient characteristics for the 19 patients used in the correlation analysis

    Discussion

    The developed method was successfully validated according to FDA guidelines.7 Only the stability of frozen patient plasma samples at 2 months did not comply with the requirements. The deviation was relatively small (mean increase of 28.8%). This increased concentration of Ara-U after 2 months of storage may be reflected by a slow conversion from Ara-C to Ara-U despite the addition of THU to the samples. However, the loss of Ara-C in the same samples was small and the divergence is more likely to be an artefact of analysis and sample handling. Two of the five samples had concentrations near the LLoQ and these two samples showed the largest difference in Ara-U concentrations over time (152.5% and 150.7%, respectively). Therefore, the non-compliance with the limits stated in the FDA guidelines probably did not influence the results of the present study. Furthermore, most of the patient plasma samples were stored for a much shorter time than 2 months before analysis.

    Only one phenotype for the deamination of Ara-C to Ara-U could be determined. The Ara-U/Ara-C ratios were similar to the ratios obtained by Burk et al3 but lower than the values found by Kreis et al2 and DeAngelis et al.11 The latter study only reported the mean values and did not analyse for different phenotypes. The existence of two distinct groups of metabolisers could have had a clinical impact, and could have led to treatment individualisation. One of the major problems with the treatment with Ara-C is drug resistance. This may be associated with irregularities in transport mechanisms, but also with induced metabolism.12 Possible ways of changing the treatment would have been to adjust the dose of Ara-C according to metaboliser group or to co-administer an inhibitor of the deaminase enzyme. The latter has previously been tried by adding THU to the treatment. The resulting mean Ara-U/Ara-C ratio from treatment with high-dose Ara-C (1.2–1.6 g/m2) and THU (2.1–2.8 g/m2) was 0.7 in a trial of eight adult patients with refractory AML and chronic myeloid leukaemia.13 This value was lower than the ratios in both the present study (median ratio 2.2, lowest ratio 0.3) and the study by Burk et al (lowest ratio 1.74). However, the difference was small, and Marsh et al also reported an unacceptable toxicity from their treatment.13 The more toxicity that follows co-administration of THU is logical since extended exposure to the cytotoxic form of Ara-C would affect the toxicity as well as the effect. This issue has also been investigated in a recent study with gemcitabine where the cytidine deaminase activity was associated with the risk of developing severe toxicity.14 Although the proposed grouping of slow and fast metabolisers was not evident from the present study, it would be of interest to investigate correlations between enzyme activity and effect or toxicity in the treatment with Ara-C for AML as well.

    Conclusions

    A division into slow and fast Ara-C metabolisers in this patient population could not be made and specific dose individualisations can therefore not be recommended. This conclusion was based on an infusion regimen with low doses (100 mg/m2) and fast infusion times (5 min) where no issues with enzyme saturations or transport mechanisms were expected, and supports the findings of Burk et al.3

    What this paper adds

    What is already known on this subject

    • Cytarabine has been used in the treatment of acute myeloid leukaemia for decades.1

    • It is rapidly deaminated in the plasma by the enzyme deoxycytidine deaminase to the inactive metabolite uracil arabinoside.

    • A previous study has outlined the existence of two phenotypes for the deamination process and suggested a link between phenotype and patient outcome.

    What this study adds

    • A population-based pharmacokinetic-pharmacodynamic method analysing dose individualisation using a selective and sensitive UV/HPLC analysis for bioanalysis of the cytotoxic cytarabine showed no phenotypic difference in the metabolism to its main metabolite.

    • Division into slow and fast Ara-C metabolisers in this patient population could not be made and specific dose individualisations can therefore not be recommended.

    References

    1. ↵
      1. Burnett A,
      2. Wetzler M,
      3. Lowenberg B
      . Therapeutic advances in acute myeloid leukemia. J Clin Oncol 2011;29:487–94.
      OpenUrlAbstract/FREE Full Text
    2. ↵
      1. Kreis W,
      2. Lesser M,
      3. Budman DR,
      4. et al
      . Phenotypic analysis of 1-B-D-arabinofuranosylcytosine deamination in patients treated with high doses and correlation with response. Cancer Chemother Pharmacol 1992;30:126–30.
      OpenUrlCrossRefPubMed
    3. ↵
      1. Burk M,
      2. Heyll A,
      3. Arning M,
      4. et al
      . Pharmacokinetics of high-dose cytarabine and its deamination product—a reappraisal. Leuk Lymphoma 1997;27:321–7.
      OpenUrlPubMedWeb of Science
    4. ↵
      1. Jamieson GP,
      2. Snook MB,
      3. Wiley JS
      . Saturation of intracellular cytosine arabinoside triphosphate accumulation in human leukemic blast cells. Leuk Res 1990;14:475–9.
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Lamba JK
      . Genetic factors influencing cytarabine therapy. Pharmacogenomics 2009;10:1657–74.
      OpenUrlCrossRefPubMedWeb of Science
    6. ↵
      1. Ueda T,
      2. Matsuyama S,
      3. Yamauchi T,
      4. et al
      . Clinical pharmacology of intermediate and low-dose cytosine arabinoside (Ara-C) therapy in patients with hematologic malignancies. Adv Exp Med Biol 1998;431:647–51.
      OpenUrlPubMed
    7. ↵
      U.S. Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research (CDER), Center for Veterinary Medicine (CVM). Guidance for Industry—Bioanalytical Method Validation. US Department of Health and Human Services, Food and Drug Administration, Center for Drug Evaluation and Research 2001. http://www.fda.gov/downloads/Drugs/.../Guidances/ucm070107.pdf (accessed 22 Feb 2013).
    8. ↵
      1. Camiener GW
      . Studies of the enzymatic deamination of ara-cytidine. V. Inhibition in vitro and in vivo by tetrahydrouridine and other reduced pyrimidine nucleosides. Biochem Pharmacol 1968;17:1981–91.
      OpenUrlCrossRefPubMedWeb of Science
    9. ↵
      1. Shapiro SS,
      2. Wilk MB
      . An analysis of variance test for normality: complete samples. Biometrika 1965;52:591–611.
      OpenUrlFREE Full Text
    10. ↵
      1. Krogh-Madsen M,
      2. Hansen SH,
      3. Honore PH
      . Simultaneous determination of cytosine arabinoside, daunorubicin and etoposide in human plasma. J Chromatogr B Analyt Technol Biomed Life Sci 2010;878:1967–72.
      OpenUrlPubMed
    11. ↵
      1. DeAngelis LM,
      2. Kreis W,
      3. Chan K,
      4. et al
      . Pharmacokinetics of ara-C and ara-U in plasma and CSF after high-dose administration of cytosine arabinoside. Cancer Chemother Pharmacol 1992;29:173–7.
      OpenUrlCrossRefPubMed
    12. ↵
      1. Cros E,
      2. Jordheim L,
      3. Dumontet C,
      4. et al
      . Problems related to resistance to cytarabine in acute myeloid leukemia. Leuk Lymphoma 2004;45:1123–32.
      OpenUrlCrossRefPubMedWeb of Science
    13. ↵
      1. Marsh JH,
      2. Kreis W,
      3. Barile B,
      4. et al
      . Therapy of refractory/relapsed acute myeloid leukemia and blast crisis of chronic myeloid leukemia with the combination of cytosine arabinoside, tetrahydrouridine, and carboplatin. Cancer Chemother Pharmacol 1993;31:481–4.
      OpenUrlCrossRefPubMed
    14. ↵
      1. Ciccolini J,
      2. Dahan L,
      3. Andre N,
      4. et al
      . Cytidine deaminase residual activity in serum is a predictive marker of early severe toxicities in adults after gemcitabine-based chemotherapies. J Clin Oncol 2010;28:160–5.
      OpenUrlAbstract/FREE Full Text

    Footnotes

    • Contributors MKM and PHH supervised and planned the study. HSJ performed the drug analysis and pharmacokinetic calculations. SHH was in charge of the bioanalytical methods. OJN and MKJ were responsible for patient care.

    • Competing interests None.

    • Ethics approval The project was approved by the Danish National Ethics committee (Journal no. H-A-2008-129).

    • Patient consent All patients signed informed consent forms before entering the trial.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Correction notice This article has been corrected since it was first published Online First. The ‘Faculty of Health and Medicinal Sciences’ had been corrected to ‘Faculty of Health and Medical Sciences’ in affiliations 1, 2 and the correspondence address.