TyrosIne Kinase Inhibitors in Chronic Myeloid Leukemia: Efficacy and Tolerability. The TIKlet Study

1. Patients

1.1. Patients

CML patients will be included according to the following inclusion criteria: a) patients of both sexes, b) age between 18 and 80 years, c) treated with imatinib or nilotinib, d) included in the follow-up in the Divisions / Units of Hematology involved in the project e) able to give informed consent to trial participation, f) with a proved compliance with the scheduled treatment.

Patients will be excluded from participation to the study if: a) age <18 or >80 years or b) unable to provide informed consent. The inability to attend the follow-up visits will not be considered an exclusion criterion from the study. In this case, collected data will be used in pharmacokinetic and statistical analyses on the basis of an intention-to-treat criterion.

The following conditions should be noted:

1. the concomitant administration of other drugs will be allowed, provided it will be known the active ingredient, the dose and the period of administration;
2. smoking and herbal products will be allowed but they should be reported on patient's case report form;
3. alterations in liver and kidney functions, body mass index higher than 28, or any other alteration in physicochemical exams do not represent exclusion criteria.

1.2. Sample size

For the purposes of this study it has been estimated that it will need to recruit at least 206 subjects per drug (total, 412 patients), taking into account:

1. the minor allele frequency of the most studied polymorphism evaluated in this research protocol (i.e., 0.3, T allele of the ABCB1 c.3435C>T polymorphism);
2. the 1:1 ratio allele C carriers and patients homozygous for the T allele of the ABCB1 c.3435C>T polymorphism;
3. covariate effects on drug pharmacokinetics will be considered significant when the mean difference in pharmacokinetic parameters will be at least 20%;
4. the power of the study will be 80%;
5. the alpha error will be set at 0.05. 1.3. Enrollment Patient enrollment in the various participating centers will be competitive, up to the achievement of 206 individuals per drug. At the first visit, the TIKlet study will be presented to patients who will satisfy the inclusion / exclusion criteria. The following activities will take place:

• patients will be informed about the characteristics, objectives, and procedures of the study.
• Consent to study participation will be confirmed by signing the informed consent form.
• Personal alphanumeric code assignment for patient's identification and privacy.
• Updating of the enrollment list.
• Collection of main data of the patient (age, sex, age at diagnosis, clinical chemistry test results as reported in the patient's record and any combination therapies, with their doses and duration) within the individual case report form (CRF).
• A blood withdrawal (approximately 4 mL) will be performed from a peripheral vein of the forearm. The time at which blood will be withdrawn will be registered within patient's CRF, together with the time elapsed from the last administration of imatinib or nilotinib. For sample handling see section 2.1. "Blood sample handling".
• The updated list of enrolled patients and individual case report forms will be sent to the Division of Hematology, Dept. of Clinical and Experimental Medicine, University of Pisa 1.4 Follow-up visits According to clinical routine at each participating center, the following activities should be performed during each follow-up visit:
• Collection of patient's data: clinical chemistry test results, cytogenetic analysis and molecular biology results (BCR/Abl transcript levels), any concomitant therapy, any adverse drug reaction.
• A blood sample (approximately 4 mL) will be collected from a peripheral vein of the forearm. The time at which blood will be withdrawn will be registered within patient's CRF, together with the time elapsed from the intake of last imatinib or nilotinib dose. For sample handling see section 2.1. "Blood sample handling".

1. Laboratory analyses Laboratory analyses will be conducted at the Division of Pharmacology, Department of Clinical and Experimental Medicine, University of Pisa (measurement of drug plasma concentrations, pharmacogenetic analyses) and at the Department of Pharmacology, University of Bologna (pharmacogenetic analyses). 2.1. Blood sample handling Each tube containing the patient's whole blood, plasma or DNA (see below) and used for the execution of therapeutic monitoring or pharmacogenetic analyses will be identified by a) the patient's code and b) the date of execution of blood withdrawal. All blood samples will be collected in lithium-heparin Vacutainer® tubes, then they will be stored at 4 °C until centrifugation. The sample will be centrifuged and the resulting plasma (1.5 mL) will be collected and stored in a tube for therapeutic drug monitoring. Only during the enrollment visit, an aliquot of whole blood (0.5 mL) will be collected in a sterile eppendorf for molecular analyses. Both samples will be stored at -20 °C for a maximum of 2 weeks, then they will be sent to the Dept. of Clinical and Experimental Medicine, University of Pisa. For sample storage see section 2.7. "Biological sample storage". 2.2 Measurement of plasma concentration of drugs The measurement of plasma concentrations of imatinib and its active metabolite, and nilotinib will be conducted by a high-performance liquid chromatography (HPLC) method using commercially available kits (Chromsystems GmbH, Munich, Germany) on Waters Breeze and Alliance instruments (Waters, USA). Briefly, solid-phase extraction columns will be balanced with Equilibration buffers 1 and 2, then the internal standard and plasma (0.5 mL) will be added. The column will be centrifuged, and subsequently washed with Wash buffer and water for HPLC. The eluate will be collected by centrifugation, diluted with water for HPLC and injected (0.025 mL) into the HPLC system for the execution of the analysis. The peak area of the analytes of interest (imatinib and its active metabolite, nilotinib and internal standard) will be recorded and the actual concentration of drug in plasma will be obtained by comparison with standard calibration curves for each day of analysis. The HPLC analyses will be carried out on a weekly basis. The reports of the analysis will be sent to the Divisions / Units of Hematology and registered into both the patient's medical record and study CRFs. 2.3. Pharmacokinetic analyses Plasma concentrations of imatinib/active metabolite and nilotinib will be analyzed trough a population pharmacokinetic analysis according to a nonlinear mixed-effects modeling approach, using the NONMEM software, version 7.2. The Xpose and PsN packages will be used to check database, for model diagnostic and covariate testing. Former models for imatinib/metabolite and nilotinib will be 1- and 2-compartment models with first-order elimination and parameterized in terms of apparent drug clearance (CL/F), volume of distribution (V/F) and absorption constant (ka), while the residual error will be modeled as additive, proportional or mixed (additive and proportional) error. The availability of records, goodness-of-fit plots and the precision of parameter estimates, together with a generalized additive modeling procedure (GAM) implemented in the Xpose package, will be adopted for model development and to explore correlations among the covariates. In particular, age, weight, body mass index, liver (ALT, AST) and renal function (creatinine plasma concentration), albumin and α1-acid glycoprotein plasma concentrations and results of pharmacogenetic analyses will be considered as possible covariates. A stepwise covariate model building will be performed with forward inclusion and backward exclusion. In particular, a decrease in the objective function value (OFV) greater than 3.84 points (i.e., p<0.05) and 6.63 points (i.e., p<0.01) will be the criteria used in the forward inclusion and backward exclusion steps, respectively. The goodness of the final model will be checked by bootstrap analysis using the PsN and Xpose packages and calculating η-shrinkage. The area under the plasma concentration-time curve from time zero up to infinity (AUC) and terminal elimination half-life (t1/2) will be calculated from the individual empirical Bayes estimates (EBE) from the final population pharmacokinetic model as follows: AUC=(Total daily dose)/(CL/F) t1/2=ln(2)/kel where kel is imatinib elimination constant. 2.4. Pharmacogenetic analyses The aliquot of frozen whole blood will be employed for the extraction of genomic DNA with a suitable kit (Qiagen, Milan, Italy) following manufacturer's instructions and the nucleic acid will be stored at -80 ° C until the time of analysis in sterile tube bearing the identification code of the patient. Single nucleotide polymorphisms (SNPs) of the following genes will be assayed: ABCB1, ABCG2, hOCT1, OCTN1, OATP1A2. SNPs will be evaluated by specific kits from Applied Biosystems (Life Sciences, Milan, Italy) on an ABI Prism 7900HT Sequence Detection System (Life Sciences). Each analysis will be performed in triplicate. Genotypes will be automatically identified through the software SDS (Life Sciences). Frequency of alleles and genotypes according to the Hardy-Weinberg law, haplotype and linkage disequilibrium analysis will be performed using Arlequin software package. 2.5. Treatment effectiveness and tolerability The response to the therapy with imatinib will be evaluated in agreement with criteria recently published and adopted. In particular, results from chemical, biochemical and molecular analyses will be performed every 2 weeks during the first 3 months of treatment, then cytogenetic (metaphase chromosomal banding assay) and molecular analyses (quantification of BCR-ABL transcription levels in plasma by real-time-PCR) were performed every 6 and 3 months, respectively, until the attainment of a complete response. The time to complete cytogenetic response (CCyR) and to major molecular response (MMR) will be defined as the length of time elapsed from diagnosis and response attainment for every patient. Furthermore, treatment-associated adverse reactions (i.e., nausea and vomiting, periorbital oedema, cramps and myalgia, neutropenia, skin and liver toxicities) will be identified and scored according to the Common Toxicity Criteria-NCI grading system (version 3). 2.6. Statistical analysis The results will be analyzed in aggregate form. Parameter of dispersion (standard deviation, range), central indexes (mean, median) and distribution (i.e., Kolmogorov-Smirnov test) will be calculated. The most appropriate statistical tests (ANOVA, Fisher test, etc.) will be applied to investigate any possible difference between groups, and the level of significance will be set at p = 0.05. The identification of cut-off values for parameters predictive of efficacy / tolerability (plasma concentrations, polymorphisms) will be pursued trough ad hoc analyses (receiver operating characteristics analysis, evaluation of positive and negative predictive values). 2.7. Biological sample storage All biological samples [whole blood (0.5 mL), plasma (1.5 mL) and genomic DNA] will be stored at the Division of Pharmacology, Department of Clinical and Experimental Medicine, University of Pisa. Aliquots of DNA samples for pharmacogenetic analyses will be stored at the Department of Pharmacology, University of Bologna. Each sample will be identified by the individual alphanumeric code assigned to every patient. According to protocol, patients may require the destruction of their own biological samples. 2.8. Authorship The criteria declared by the ICMJE (International Committee of Medical Journal Editors) will be adopted for authorship definition.