Cancer Therapy: Clinical

Effect of Common CYP3A4 and CYP3A5 Variants on the Pharmacokinetics of the 3A Phenotyping Probe Midazolam in Cancer Patients Erin R. Lepper,1 Sharyn D. Baker,5 Matt Permenter,2 Nicole Ries,5 Ron H.N. van Schaik,7 Paul W. Schenk,7 Douglas K. Price,2 Danielle Ahn,2 Nicola F. Smith,2 George Cusatis,5 Roxann G. Ingersoll,6 Susan E. Bates,3 Ron H.J. Mathijssen,8 Jaap Verweij,8 William D. Figg,2,4 and Alex Sparreboom4

Abstract Purpose:To evaluate the effect of naturally occurring variants in encoding the cytochrome P450 (CYP) isoforms CYP3A4 and CYP3A5 in patients with cancer receiving midazolam as a phenotyping probe. Experimental Design: Five variants in CYP3A4 and CYP3A5 were evaluated in 58 patients (21women and 37 men) receiving a short i.v. bolus of midazolam (dose, 0.0145 or 0.025 mg/kg). Midazolam concentrations in plasma were determined using liquid chromatography-mass spec- trometry, and pharmacokinetic variables were calculated using noncompartmental analysis. Genomic DNA was characterized for the variants by PCR-RFLP, and all genotypes were con- firmed by direct nucleotide sequencing. Results:The mean clearance of midazolam was 24.4 F 9.12 L/h, and phenotypic CYP3A activity varied about 4-fold in this population (range, 10.8-44.3 L/h). There were six carriers of the CYP3A4*1B allele (allele frequency, 0.061). No variant alleles for CYP3A4*17, CYP3A4*18A , or CYP3A5*6 were identified. Forty-eight of the 58 patients were homozygous variant for CYP3A5*3C, eight were heterozygous, and two were homozygous wild type (allele frequency, 0.897). No associations were noted between any of the studied genotypes and the phenotypic measures (P z 0.16).Likewise, a common variant in exon 26 in the encoding P-glycoprotein [i.e., ABCB1 (MDR1) 3435C>T] that was previously reported to be linked to CYP3A4 mRNA levels was unrelated to any of the studied phenotypic measures (P z 0.49). Conclusions.The studied genetic variants in CYP3A4 and CYP3A5 are unlikely to have an impor- tant functional significance to phenotypic CYP3A activity in patients with cancer.

Isoforms of the cytochrome P450 (CYP) 3A subfamily are the (1). In adults, CYP3A4 and CYP3A5 are predominant among most abundantly expressed CYP in the human liver the four known isoforms (CYP3A4, CYP3A5, CYP3A7, and CYP3A43) in the liver and intestines (1). However, metabol- ically active CYP3A5 is expressed only by an estimated 10% to Authors’ Affiliations: 1Science Applications International Corporation-Frederick, 30% of Caucasians and 50% to 60% of African Americans 2Molecular Pharmacology Section, 3Cancer Therapeutics Branch, 4Clinical (2, 3). The genetic basis for polymorphically expressed CYP3A5 5 Pharmacology Research Core, National Cancer Institute, Bethesda, Maryland; The was recently reported to be associated with a single nucleotide Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, 6Institute of Genetic Medicine,The Johns Hopkins University School of Medicine, Baltimore, polymorphism (SNP) in intron 3 of the CYP3A5 gene Maryland; Departments of 7Clinical Chemistry, Erasmus MC and 8Medical (CYP3A5*3C; refs. 4, 5). This SNP causes alternative splicing Oncology,Erasmus MC-Danielden Hoed Cancer Center, Rotterdam, the Netherlands and truncation of CYP3A5 and provides the basis for Received 3/9/05; revised 5/10/05; accepted 5/18/05. the absence of CYP3A5 protein in most individuals of Grant support: National Cancer Institute contract N01-CO-12400 (E.R. Lepper) and the CornelisVrolijk Development Fund, IJmuiden, the Netherlands (J.Verweij). Caucasian descent. In people with at least one CYP3A5*1A The costs of publication of this article were defrayed in part by the payment of page wild-type allele, CYP3A5 accounts for at least 50% of the total charges. This article must therefore be hereby marked advertisement in accordance CYP3A content, and results in f2- to 3-fold higher total CYP3A with 18 U.S.C. Section 1734 solely to indicate this fact. activity in vitro (4, 6). The most prevalent polymorphism in Disclaimer: The content of this publication does not necessarily reflect the views or CYP3A4 (CYP3A4*1B) occurs in the 5V flanking region of the policies of the Department of Health and Human Services nor does mention of trade names, commercial products, or organization imply endorsement by the U.S. gene. The frequency of this allele ranges from 2% to 9.6% in Government. Caucasians to 35% to 67% in African Americans (1). Requests for reprints: Alex Sparreboom, Clinical Pharmacology Research Core, A wide variation in CYP3A expression and activity exists in National Cancer Institute, Room 5A01, Building 10, 9000 Rockville Pike, Bethesda, humans, which is reflected by extensive interindividual differ- MD 20892. Phone: 301-402-9498; Fax: 301-402-8606; E-mail: SparrebA@ mail.nih.gov. ences in the clearance of CYP3A substrate drugs and phenotyp- F 2005 American Association for Cancer Research. ing probes like midazolam (7). For many anticancer drugs used doi:10.1158/1078-0432.CCR-05-0520 therapeutically, pharmacokinetic variability is a significant

ClinCancerRes2005;11(20)October15,2005 7398 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 24, 2021. © 2005 American Association for Cancer Research. CYP3A Phenotype/Genotype Relations contributor to the variability of both the severity of side effects reaction volume of 20 AL. The temperature profile for the PCR reaction and therapeutic response (8). Furthermore, the metabolism of was one cycle at 94jC for 5 minutes followed by 20 cycles of 94jC for j j 37% of all currently approved cytostatic and/or cytotoxic 30 seconds, 62 C for 30 seconds, and 72 C for 30 seconds, with a final j anticancer agents is known to be mediated, at least in part, by 7-minute cycle at 72 C. After initial amplification, 3 AL of amplified product were removed and an additional PCR amplification was done the CYP3A isoforms. Hence, identification of factors affecting using 40 pmol of the primers 5V-GCTCTGTCTGTCTGGGTTTGG-3V the clearance of CYP3A substrates could aid in predicting or (CYP3A4*1B F4) and 5V-CACACCACTCACTGACCTCCT-3V adapting appropriate, individualized doses of anticancer drugs. (CYP3A4*1B R4) in a 50-AL reaction. PCR conditions were 94jC for Here, we evaluated the effect of common naturally occurring 5 minutes followed by 40 cycles of 94jC for 30 seconds, 64jC for variants in genes encoding CYP3A4 and CYP3A5 in cancer 30 seconds, and 72jC for 30 seconds, with a final 7-minute cycle at 72jC. patients receiving midazolam as a phenotyping probe. Sequencing was done using the following primers: (CYP3A4*1B F5) 5V-AGGTGTGGCTTGTTGGGATG-3V for the forward strand and (CYP3A4*1B R5) 5V-TCAGAAACTCAAGTGGAGCC-3V for the reverse Materials and Methods strand. For the CYP3A5*3C allele, primers were designed and chosen as Patient selection. This study was conducted under two Institutional follows: 5V-TCCTCAGAATCCACAGCGCTG-3V(CYP3A5*3C F1) and 5V- Review Board–approved protocols and all patients signed an informed TTTATGTGCTGGAGAAGGACG-3V (CYP3A5*3C R1). These primers consent form before treatment. Patients with a histologically or were used in a 20-AL reaction that contained 1 PCR buffer (Perkin- cytologically confirmed diagnosis of malignant solid tumor were eligible Elmer), 1.5 mmol/L MgCl2, 2 mmol/L deoxynucleotide triphosphates, for two clinical trials in which patients underwent CYP3A phenotyping and 2 units of Platinum Taq polymerase (Invitrogen). The cycle with i.v. given midazolam (9, 10). Eligibility criteria included age >18 conditions were: one cycle at 94jC for 5 minutes followed by 20 cycles years; performance status V1; and adequate hematopoietic (leukocytes, of 94jC for 30 seconds, 64jC for 30 seconds, and 72jC for 30 seconds, z4.0 109/L; neutrophils, z1.5 109/L; and platelets, z100 109/L), ending with a final 7-minute cycle at 72jC. After initial amplification, hepatic (bilirubin within normal limits; transaminases, V2times 3 AL of amplified product were removed and an additional PCR the upper limit of normal), and renal function (creatinine clearance, amplification (as described above in a total of 50 AL final volume) was z50 mL/min). All patients were required to have an estimated life done using 40 pmol of primers 5V-AGCACTTGATGATTTACCTGCC-3V expectancy of >12 weeks and no previous chemotherapy was allowed for (CYP3A5*3C F2) and 5V-CCAGGAAGCCAGACTTTGATC-3V at least 4 weeks before enrollment. The patients were asked to abstain (CYP3A5*3C R2). PCR conditions for this reaction were 94jC for from substances known to affect the function and/or expression of 5 minutes followed by 40 cycles of 94jC for 30 seconds, 62jC for CYP3A and/or ABCB1 for a period of 2 weeks before, during, and up to 30 seconds, and 72jC for 30 seconds, with a final 7-minute cycle at 72jC. 3 weeks after the administration of midazolam. Sequencing primers were as follows: (CYP3A5*3C F3) 5V-AGTGGCA- Blood sampling and processing. Blood samples were collected in glass TAGGAGATACCCAC-3V and (CYP3A5*3C R3) 5V-AGGTTCTAGTTCAT- tubes containing heparin. Samples were drawn at 5 and 30 minutes and TAGGGTG-3V. 1, 2, 3, 4, and 5 hours following bolus i.v. administration of midazolam For the CYP3A5*6 allele, an initial PCR was done using the sample at a dose of either 0.0145 or 0.025 mg/kg. In about half of the patients, DNA and 20 pmol of each primer, 5V-AGGGAGTAGATGGAAGAT- additional samples were obtained at 15 minutes and 6 hours after GATTC-3V(CYP3A5*6 F1) and 5V-AGTTGATTATTGGATGCTTAGGGC-3V midazolam administration. Immediately after collection, each sample (CYP3A5*6 R1). These primers were used in the initial reaction along was centrifuged for 10 minutes at 2,000 g (4jC), and plasma with 1 PCR buffer (Perkin-Elmer), 1.5 mmol/L MgCl2, 2 mmol/L supernatants were stored at 80jC until the day of analysis. deoxynucleotide triphosphates, and 2 units of Platinum Taq polymer- Drug measurement and pharmacokinetic analysis. Quantitation of ase (Invitrogen) in a reaction volume of 20 AL. The temperature profile midazolam in plasma was done by high-performance liquid chroma- for the PCR reaction was one cycle at 94jC for 5 minutes, 40 cycles of tography with mass spectrometric detection, as described previously 94jC for 30 seconds, 64jC for 30 seconds, and 72jC for 30 seconds (11). Pharmacokinetic profiles of midazolam were analyzed by non- followed by a final 7-minute cycle at 72jC. After initial amplification, compartmental methods using the software package WinNonlin version 3 AL of amplified product were removed and an additional PCR 4.0 (Pharsight, Mountain View, CA). The variables calculated included amplification using the above reaction ingredients was done using area under the plasma concentration-time curve extrapolated to infinity 40 pmol of the designed primers 5V-TGCTGCATGTATAGTGGAAGGAC-3V (AUC), clearance (CL, calculated as dose/AUC), volume of distribution

at steady-state (Vss), and half-life of the terminal phase (t1/2, z). The AUC was normalized to a dose of 0.025 mg/kg. As a comparison, the AUC was also calculated based on the limited sampling model Ta b l e 1. Demographic characteristics described previously (12). CYP3A4 and CYP3A5 genotype analysis. Plasma was used to Patient demographics isolate genomic DNA according to the manufacturer’s instructions Gender Female 27 using the UltraSens Virus Kit (Qiagen, Valencia, CA). Variations in Male 31 CYP3A4 (CYP3A4*1B, CYP3A4*17, and CYP3A4*18A) and CYP3A5 Race Caucasian 55 (CYP3A5*3C and CYP3A5*6) were analyzed using RFLP-based techniques, as previously described (13, 14). Confirmation of all African American 2 the variant genotype assignments (i.e., for CYP3A4*1B, CYP3A5*3C, Hispanic 1 and CYP3A5*6) was done using direct nucleotide sequencing. In preparation for sequencing, an initial PCR was done with Mean Median Min Max concentrated DNA using 20 pmol of each primer. For CYP3A4*1B, the primers were designed and chosen as follows: 5V-CTGTGTGAG- Age (y) 56.7 58 29 73 GAGTTTGGTGAG-3V (CYP3A4*1B F3) and 5V-TGGAAGAGGCTTCTC- Height (m) 1.73 1.74 1.55 1.91 CACCTTG-3V(CYP3A4*1B R3). These primers were added to a reaction Weight (kg) 80.27 77.85 48 137 mixture of 1 PCR buffer (Perkin-Elmer, Norwalk, CT), 2 mmol/L of BSA (m2) 1.94 1.91 1.46 2.46 each of the four deoxynucleotide triphosphates, 1.5 mmol/L MgCl2, Bilirubin (Ag/dl) 0.50 0.47 0.23 1.23 and 2 units of Platinum Taq polymerase (Invitrogen, Carlsbad, CA) in a

www.aacrjournals.org 7399 Clin Cancer Res 2005;11(20) October 15, 2005 Downloaded from clincancerres.aacrjournals.org on September 24, 2021. © 2005 American Association for Cancer Research. Cancer Therapy: Clinical

Ta b l e 2 . Summary of midazolam pharmacokinetics

Variable Mean Median SD Min Max Sample size Population (route) AUC (ng h/mL) 90.2 80.6 35.8 39.0 204

Cmax (ng /m L) 79 . 2 71. 2 31. 9 3 7. 5 19 4

Vss (L) 78.0 70.1 40.1 24.9 228

T1/2 , z (h) 3.40 2.8 2.37 0.62 17.5 CL (L/h) This study 24.4 22.8 9.12 10.8 44.3 58 Cancer patients (i.v.) (15) 25.6 * 9.88 * * 31 Cancer patients (i.v.) c (28) 34.9 ****22CYP3A5*3C/3C volunteers (orally) c 31.9 ****17CYP3A5*1A/*3C volunteers (orally) (42) 25.4 * 9.3 * * 27 Cancer patients (i.v.) 24.4 * 8.1 * * 24 Volunteers (i.v.) (25) 23.2 * 6.3 9 47.6 23 Volunteers (i.v.) c c (24) 112 12 3 * 55.6 214 21 Volunteers (orally) (16) 26.9 * 12.8 8.7 68.7 45 Cancer patients (i.v.) 81.7 * 41.5 7.74 185 22 Cancer patients (orally) (29) 32.5 * 10.2 * * 6 CYP3A5*1A/*1A volunteers (i.v.) 32.1 * 14.5 * * 6 CYP3A5*1A/*3C volunteers (i.v.) 28.4 * 12.5 * * 7 CYP3A5*3C/*3C volunteers (i.v.)

Abbreviation: Cmax, dose-normalized peak plasma concentration. *Data not provided. cCalculated based on AUC and dose published.

(CYP3A5*6 F2) and 5V-TTTGTGGTGGGGTGTTGACAGC-3V(CYP3A5*6 sequence analysis were used for correlation with midazolam pharma- R2). PCR conditions for this reaction were 94jC for 5 minutes followed cokinetic variables. by 40 cycles of 94jC for 30 seconds, 66jC for 30 seconds, and 72jCfor ABCB1 (MDR1) genotype analysis. The following primers were 30 seconds, with a final 7-minute cycle at 72jC. The sequencing primers designed for the ABCB1 3435C>T variant in exon 26: 5V-AGCAACCT- were 5V-TTGGGGCCTACAGCATGGATG-3V (CYP3A5*6 F3) and 5V- TACATCTACTAC-3V (ABCB1 3435C>T 6 F) and 5V-GGAGAGACAGT- TGTGTGAGGGCTCTAGATTGAC-3V (CYP3A5*6 R3). The existence of CATGCCTAC-3V (ABCB1 3435C>T R) and nested primers each variant was determined by direct nucleotide sequencing using the 5V-TCACAGTAACTTGGCAGTTTCA-3V (ABCB1 3435C>T F) and 5V- dRhodamine Terminator Cycle Sequencing Ready Reaction kit on an GGGACCAGCCCCTTATAAATC-3V (ABCB1 3435C>T R; Invitrogen). ABI 310 genetic analyzer (Applied Biosystems, Foster City, CA). The outside primers and f100 ng of genomic DNA were added to a Concordance in results between RFLP and direct sequence analysis 15-AL reaction mixture consisting of 1 mmol/L deoxynucleotide was observed in all but one sample (for CYP3A5*3C only); results from triphosphates, 1 PCR buffer, 2.5 mmol/L MgCl2, and 0.6 units of

Ta b l e 3 . Genotype and allele frequencies for the studied variants

c b Polymorphism* Nomenclature Effect Genotype frequencies Allele frequenciesx Wt Het Var p q

CYP3A4 392A>G CYP3A4*1B Promoter 51 (89.5) 5 (8.77) 1 (1.75) 0.939 0.061 CYP3A4 15 615T>C CYP3A4*17 F189S 58 (100) 0 (0) 0 (0) 1.000 0.000 CYP3A4 20070T>C CYP3A4*18A L293P 58(100) 0(0) 0(0) 1.000 0.000 CYP3A5 6986A>G CYP3A5*3C Splicing defect 2 (3.45) 8 (13.79) 48 (82.8) 0.103 0.897 CYP3A5 146 9 0G>A CYP3A5*6 Splicing defect 58 (100) 0 (0) (0) 1.000 0.000 ABCB1 3435C>T Not available I1145I 12(20.69) 35 (60.34) 11 (18.97) 0.509 0.491

Abbreviations: Wt, homozygous wild-type patient; Het, heterozygous variant patient; Var, homozygous variant patient. *Number represents position in nucleotide sequence based on gene AF280107 (for CYP3A4*1B, CYP3A4*17,andCYP3A4*18A). There is still no genomic CYP3A5 sequence available corresponding to the CYP3A5*1A allele initially described on the cDNA level and encoding a functional CYP3A5 .The reference sequence for the CYP3A5*1A allele has been obtained by using the CYP3A5 sequence found in accession no NG___000004.2 (corresponding to the CYP3A5*3A allele) and replacing base 6986 with an A and base 31611with a C (numbering based on translation start as +1). See http://www.imm.ki.se/CYPalleles/cyp3a4.htm and http://www.imm. ki.se/CYPalleles/cyp3a5.htm). cNumber represents amino acid codon. bNumber represents number of patients with percentage in parenthesis; the difference in total number of patients for CYP3A4*1B is due to the fact that one sample did not shown PCR amplification. xHardy-Weinberg notation for allele frequencies (p, frequency for wild-type allele and q, frequency for variant allele).

ClinCancerRes2005;11(20)October15,2005 7400 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 24, 2021. © 2005 American Association for Cancer Research. CYP3A Phenotype/Genotype Relations

Fig. 1. Midazolam clearance as a function of the CYP3A4*1B, CYP3A5*3C,andABCB1 3435C>T genotypes. Horizontal lines indicate mean values in each group. Abbre- viations: Wt, homozygous wild-type patient; Het, heterozygous patient; Var, homozygous variant patient. Two-sided Ps are given above the panels.

Amplitaq Gold (Applied Biosystems). For the nested PCR reaction, mainly exploratory in intent, a Bonferroni adjustment was used to 2 AL of the first PCR product were used as template for amplification. evaluate the significance of the multiple comparisons. Two-sided Ps< The thermocycler conditions were an initial cycle at 95jC for 12 0.017 (0.05 divided by the three observed variant genotypes) were minutes followed by 10 cycles 94jC for 45 seconds, 55jC for 1 regarded as statistically significant and Ps < 0.05 were considered a minute and 72jC for 1 minute, 20 cycles of 89jC for 1 minute, trend. These levels were chosen to reduce the risk of finding purely 55jC for 1 minute and 72jC for 1 minute, and ending with a final coincidental associations in view of the number of variant genotypes 10-minute cycle at 72jC. The restriction enzyme MboI (New England analyzed concurrently. Biolabs, Beverly, MA) was used for the determination of the variants. Enzyme (0.5 AL) and 2 AL of enzyme buffer were added to 20 ALof PCR product and incubated at 37jC for 1 hour followed by an Results incubation at 65jC for 20 minutes. Blue juice (1 ALof5 solution) was added to 10 AL restricted PCR product and run on a 2.0% Patients. A total of 58 adult patients (21 women and 37 men) agarose gel with ethidium bromide as a fluorescence detector. A ‘‘T’’ were evaluated during this study. Patients were of Caucasian mutation was determined by bands of 408, 263, 158, 103, and 59 (n = 55), African American (n = 2), or Hispanic (n = 1) descent bp; a ‘‘C’’ was determined by bands of 263, 236, 172, 158, and 59 and were between 29 and 73 years old (median age, 57 years). bp; heterozygous samples contained all bands. The existence of the Additional patient demographics are shown in Table 1. variant was also determined by direct nucleotide sequencing after a Midazolam disposition. Complete midazolam pharmacoki- cleanup step with MinElute PCR purification kit (Qiagen) in a total netic data following i.v. administration were available in all 58 volume of 12 AL, which included 6 AL of DNA at 3 ng per 100 bp of patients. The mean dose-normalized AUC was 90.2 F 35.8 ng amplified products and 6 AL of either the forward or reverse primer. h/mL, whereas the mean clearance was 24.4 F 9.12 L/h, which Sequencing in the forward and reverse direction was done using an ABI 3730XL Sequencer (Applied Biosystems). Concordance in results is consistent with earlier findings obtained in cancer patients between RFLP and direct sequence analysis was observed in all but following i.v. administration of midazolam (15, 16). Table 2 four samples; results from sequence analysis were used for correlation provides a comparison of the clearance values determined in with midazolam pharmacokinetic variables. previously conducted trials in cancer patients and healthy Statistical considerations. All data are presented as mean values F volunteers, which involved genotyping for CYP3A5*3C. SD, unless stated otherwise. Genotype-frequency analysis of Hardy- Variant genotypes. Genotyping results were available for all Weinberg equilibrium was carried out using Clump version 1.9. Linkage patients except for CYP3A4*1B, which had available data for 57 disequilibrium was calculated using EMLD (Qiqing Huang; see http:// of the 58 patients (Table 3). Of these 57 patients, there were linkage.rockefeller.edu/soft/). The linkage between each pair of SNPs six carriers of the CYP3A4*1B allele (five heterozygous was determined in terms of the classic statistic DV. The absolute value CYP3A4*1A/*1B and one homozygous CYP3A4*1B/*1B). Both for DV (|DV|) of 1 denotes complete linkage disequilibrium, whereas a value of 0 denotes complete linkage equilibrium. of the subjects of African American descent were wild type for To relate pharmacokinetic variables with the variant genotypes, the CYP3A4*1B (homozygous CYP3A4*1A/*1A). No patients were Kruskal-Wallis test, a nonparametric one-way ANOVA, was used for identified carrying variant alleles for CYP3A4*17, CYP3A4*18A, each polymorphism examined. (NCSS v2001; J. Hintze, Number or CYP3A5*6. Forty-eight of the 58 patients were homozygous Cruncher Statistical Systems, Kaysville, UT). Although this study was variants for CYP3A5*3C and there were eight heterozygotes

www.aacrjournals.org 7401 Clin Cancer Res 2005;11(20) October 15, 2005 Downloaded from clincancerres.aacrjournals.org on September 24, 2021. © 2005 American Association for Cancer Research. Cancer Therapy: Clinical

(CYP3A5*1A/*3C). The ABCB1 3435C>T genotypes were distributed as follows: 12 wild type (C/C), 35 heterozygous (C/T), and 11 variant (T/T). ABCB1 3435C>T was in linkage equilibrium with both CYP3A4*1B and CYP3A5*3C (DV = 0.183 and 0.0933, respec- tively). Linkage disequilibrium was apparent between CYP3A4*1B and CYP3A5*3C (DV = 0.656), as described previously (17). Genotype-phenotype relations. There were no significant cor- relations between any of the variant genotypes and clearance, volume of distribution at steady state, or AUC (Fig. 1; Table 4). Although the mean midazolam clearance was 1.2 times higher in patients with the CYP3A5*1A/*1A genotype compared with patients that were homozygous variant (CYP3A5*3C/*3C), this difference was not statistically significant (P = 0.50). Furthermore, there was no correlation between CYP3A5*3C genotype status and clearance, when patients with the homo- zygous variant genotype (CYP3A5*3C/*3C) were compared Fig. 2. Scatter plot of the observed AUC of midazolam versus theAUC of with patients that had at least one wild-type allele (CYP3A5*1A/ midazolam calculated based on the observed midazolam concentration at the *1A and CYP3A5*1A/*3C; P = 0.24). 4-hour sampling time point (4-h conc.) using the limited-sampling model equation: Comparison of predicted and observed area under the plasma AUC = 9.91 + 12.2 [4-h conc.], as described previously (12).The dotted line indicates a linear regression fit, whereas the solid line represents the line of identity. concentration-time curve. The AUC of midazolam was also calculated with a limited-sampling model based on the 4-hour It has been proposed that genotyping for CYP3A4 and time point, as done by Wong et al. (16). This model was designed CYP3A5 variants may be useful for prediction of total CYP3A originally using a different data set, as described elsewhere (12). activity because of the genetic diversity in the genes encoding Figure 2 shows the correlation of the observed AUC with the these (2, 3). Over 30 SNPs in CYP3A4 have been calculated AUC. The correlation between the predicted AUC and 2 published; however, most are unlikely to contribute substan- the observed AUC, using all time points, showed an R = 0.749. tially to the interindividual variability of CYP3A4 activity F The mean ( SE) bias and precision of the prediction made by in vivo, based on their limited functional significance and/or the model, determined by calculation of the percentage mean low allele frequency (3, 19, 20). Similar to earlier findings (21), prediction error and the relative root mean squared prediction the very low allele frequency of the functional CYP3A4*17 and F F error, were 18.3 5.82% and 30.2 3.73%, respectively. CYP3A4*18A variants evaluated in the present study suggests that these SNPs most likely have no broad relevance to CYP3A4 Discussion activity and function in predominantly Caucasian populations. The one additional SNP that was studied here, CYP3A4*1B,isa Interindividual variability in drug efficacy and toxicity, promoter variant in the so-called nifedipine-specific element. resulting in unpredictable patient responses, is commonly The allele frequency of this variant is known to depend on observed in all therapeutic areas (18). However, these differ- racial ancestry, being absent in Japanese and Chinese subjects, ences are particularly important in the field of cancer therapy and present in 2% to 10% in Caucasians and 35% to 67% in because many anticancer agents have a narrow therapeutic African Americans, respectively (3). The currently observed index. It is now well recognized that in addition to environ- frequency of 6.1% is consistent with previous estimates for mental and physiologic factors, such interpatient variation is predominantly Caucasian populations. Although the function- commonly associated with polymorphisms in genes encoding al significance of this SNP presently remains unclear, the drug-metabolizing enzymes, drug transporters, and/or drug preponderance of the evidence suggests that CYP3A4*1B does targets. In this study, we tested the hypothesis that the dis- not affect the metabolism and clearance of CYP3A4 substrate position of the prototypical CYP3A substrate drug midazolam drugs (22–26). Likewise, in the current study, the pharmaco- in cancer patients is dependent on genetic variability in the kinetic profile of midazolam was not significantly different CYP3A4 and CYP3A5 genes. between patients with or without the CYP3A4*1B variant allele.

Ta b l e 4 . Genotype-phenotype relationships

Polymorphism CL (L/h) Wt Het Var P Wt CYP3A4*1B 22.5 (10.8-44.3, 51) 32.1 (22.4-42.2, 5) 22.0 (1) 0.16 77.6 (24.9-228, 51) CYP3A5*3C 28.5 (16.8-40.2, 2) 26.4 (16.5-42.2, 8) 22.3 (10.8-44.3, 48) 0.50 81.0 (42.5-119, 2) ABCB1 3435C>T 22.8 (10.8-43.3,12) 23.6 (11.4-44.3, 35) 18.3 (11.4-40.8,11) 0.56 88.8 (32.8-130,12)

NOTE: Numbers represent median (range, number of subjects with each genotype) or two-sided Ps. Abbreviations: Wt, homozygous wild-type patient; Het, heterozygous variant type patient; Var, homozygous variant type patient.

ClinCancerRes2005;11(20)October15,2005 7402 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 24, 2021. © 2005 American Association for Cancer Research. CYP3A Phenotype/Genotype Relations

As mentioned previously and in contrast to CYP3A4, the previous study (16). Collectively, current evidence does not CYP3A5 protein isoform is known to be expressed in only a suggest that CYP3A5*3C-associated differences in the extent of small percentage of Caucasian individuals and this has been drug metabolism of CYP3A substrates are clinically important. linked to a common transition in intron 3 of the CYP3A5 gene The extent to which the CYP3A isoforms metabolize (CYP3A5*3C), which introduces a frameshift during transl- midazolam is likely to be determined, at least in part, by ation and results in a truncated, nonfunctional protein (4, 5). intracellular drug concentrations in the liver and, to a lesser Approximately 70% to 90% of Caucasian subjects are homozy- extent, the intestine. This process, in turn, is partially dependent gous variant for CYP3A5*3C and thus are deficient in functionally on the ATP-binding cassette transporters like ABCB1 active CYP3A5 (27), which is consistent with the currently (P-glycoprotein) that are localized in the apical membrane of observed genotype frequency of 89.7%. In the present study, no hepatocytes and enterocytes. Although it has been suggested change in midazolam pharmacokinetic variables was noted that midazolam is not a substrate for ABCB1 (32), recent data between 10 patients with at least one wild-type (CYP3A5*1A) have indicated that midazolam exhibits characteristics of a allele and 48 patients carrying two variant (CYP3A5*3C) alleles. highly permeable ABCB1 substrate (33). The clearance of This is similar to findings obtained in healthy subjects using midazolam has also been shown to be correlated (r = 0.6; P = midazolam (24, 25, 28, 29), or various other CYP3A phenotyping 0.0005) with the clearance of docetaxel (15), a known dual probes, including erythromycin (23, 25) and nifedipine (30). substrate of CYP3A and ABCB1 (34), suggesting that varying Interestingly, in a small cohort of Australian cancer patients, expression of ABCB1 may alter midazolam clearance. In recent Wong et al. recently observed f1.4-fold higher systemic clearance years, various genetic variants in the ABCB1 gene have been of midazolam in four patients with the CYP3A5*1A/*3C described that may effect transporter expression or function genotype compared with 39 patients with the CYP3A5*3C/*3C (35). The most extensively studied ABCB1 variant to date is a genotype (P = 0.01; ref. 16). In an effort to explain this discrepant common synonymous C-to-T transition at nucleotide position finding, it is noteworthy that the interindividual variability in 3435 at a wobble position in exon 26 (36). Although this CYP3A phenotypic activity was substantially greater in the study transition does not change its encoded amino acid, recent reported by Wong et al. than that found in the various other findings have indicated that this variant is associated with studies [8-fold for systemic clearance and 24-fold for apparent altered protein expression in different human tissues (37), and oral clearance (16) versus 3.3- to 5.2-fold (24, 25, 28, 29)], this may result in decreased hepatobiliary and/or intestinal suggesting the possibility of altered expression of either CYP3A4, secretion of substrate drugs. Furthermore, a reduced expression CYP3A5, or both as a result of differences in advancement of the of intestinal CYP3A4 mRNA has been observed in subjects disease state of patients in the studied cohort (31). carrying the homozygous variant genotype of the ABCB1 Wong et al. estimated the systemic midazolam clearance 3435C>T polymorphism in a Japanese population (38), further from a previously developed limited-sampling model based on pointing to the possibility of ABCB1 genotype affecting the collection of one plasma sample obtained 4 hours after midazolam clearance. However, in this study, midazolam drug administration (16). Although a retrospective validation pharmacokinetic variables were not significantly influenced of this model on our own data set indicated a good correlation by the ABCB1 3435C>T genotype, which is consistent with between the predicted AUC and the observed AUC (R2 = prior observations in healthy volunteers (25, 39). It should be 0.749), the model had a statistically significant tendency for noted that the lack of relationships with this SNP is, as systematic negative error and showed poor precision. To expected, consistent with preclinical observations in Abcb1a- illustrate this problem, two different patients in our cohort deficient mice that metabolism rather than transport is the that had very similar observed AUC values of 117.0 and prominent elimination pathway for midazolam (40). This 117.2 ng h/mL, respectively, had a predicted AUC based on the furthersuggeststhattheinvolvementofABCB1inthe limited-sampling model of 37.3 and 165 ng h/mL, respectively. disposition of midazolam may be relatively unimportant This raises the possibility that some of the proposed genotype- regardless of ABCB1 genotype status. phenotype associations described previously that were based on The lack of statistically significant relations in this study does variable estimates from a limited-sampling model might be not necessarily mean that there are none, especially in light particularly prone to the risk of false-positive error due to of the few individuals studied in our population with a inaccurate phenotypic assessments. In addition, some error homozygous variant genotype. However, the variant genotype may also be attributed to the smaller sample size used in the effects are minor and/or the genotype heterogeneity is

Ta b l e 4 . Genotype-phenotype relationships (Cont’d)

Vss (L) AUC (ng h/mL) Het Var P Wt Het Var P

49.9 (36.8-189, 5) 69.9 (1) 0.97 80.7 (39.0-204, 51) 69.0 (49.5-108, 5) 99.9 (1) 0.46 45.6 (24.9-189, 8) 73.8 (32.8-228, 48) 0.77 86.4 (60.0-113, 2) 73.2 (49.5-117, 8) 86.8 (39.0-204, 48) 0.56 69.9 (31.9-228, 35) 69.7 (24.9-145,11) 0.87 77.2 (47.8-204,12) 89.0 (39.0-168, 35) 79.3 (53.4-166,11) 0.85

www.aacrjournals.org7403 Clin Cancer Res 2005;11(20) October 15, 2005 Downloaded from clincancerres.aacrjournals.org on September 24, 2021. © 2005 American Association for Cancer Research. Cancer Therapy: Clinical sufficiently small to conclude that the presently studied variants also been postulated to influence CYP3A activity. Furthermore, in the CYP3A4 and CYP3A5 genes do not cause a substantial many drugs and herbal medications are known to interact with interindividual difference in midazolam clearance and thus are CYP3A. Although patients were asked to abstain from any drugs unlikely to have an important functional significance in cancer known to affect the function and/or expression of either CYP3A patients treated with CYP3A substrate drugs. For example, even or ABCB1, this possibility cannot be entirely excluded. Finally, it in a population of this size, if CYP3A5 genotype accounted for is possible that additional genes, genetic variants and/or the large degree of variability in midazolam pharmacokinetics, haplotypes of importance to the pharmacokinetics of CYP3A detection of this effect would have been expected. substrate drugs like midazolam may yet be discovered and/or There may be a number of other factors contributing to the evaluated. Further investigation is needed to determine the variability in midazolam pharmacokinetics, because the tested relative role of genetic variation and environmental variables genetic variants do not seem to explain this finding. A recent (e.g., concomitant drugs or herbs) on the pharmacokinetics of study has shown an inverse correlation between liver dysfunc- anticancer drugs that are substrates of CYP3A4 and CYP3A5. tion, particularly total bilirubin levels, and CYP3A activity (23). Although all participants in the current study had total bilirubin Acknowledgments levels within 1.5 times the upper limit of normal, there was still a considerable range represented (Table 1). Inflammatory re- We thank Marloes van der Werf and Ilse van der Heiden (Department of Clinical Chemistry, Erasmus MC, Rotterdam, the Netherlands) for expert technical assis- sponse in patients with cancer, resulting in increased a-1 acid tance and Dr. Seth M. Steinberg (Center for Cancer Research, Biostatistics and Data glycoprotein serum concentrations, has also been suggested to Management Section, National Cancer Institute, Bethesda, MD) for critical review of decrease clearance by the CYP3A isozymes (41). Age and sex have the article.

References 1. Lamba JK, LinYS, Schuetz EG,Thummel KE. Genetic Gurney H. CYP3A5 genotype and midazolam clear- genotype on the pharmacokinetics of intravenous contribution to variable human CYP3A-mediated ance in Australian patients receiving chemotherapy. midazolam during inhibited and induced metabolic metabolism. Adv Drug Deliv Rev 2002;54:1271 ^ 94. Clin PharmacolTher 2004;75:529^ 38. states. Clin PharmacolTher 2004;76:104 ^ 12. 2. Wojnowski L. Genetics of the variable expression of 17. Fukushima-Uesaka H, Saito Y,Watanabe H, et al. 30. Fukuda T, Onishi S, Fukuen S, et al. CYP3A5 geno- CYP3A in humans.Ther Drug Monit 2004;26:192 ^ 9. Haplotypes of CYP3A4 and their close linkage with type did not impact on nifedipine disposition inhealthy 3. Xie HG, Wood AJ, Kim RB, Stein CM,Wilkinson GR. CYP3A5 haplotypes in a Japanese population. Hum volunteers. Pharmacogenomics J 2004;4:34 ^ 9. Genetic variability in CYP3A5 and its possible conse- Mutat 2004;23:100. 31. Wilkinson GR. Genetic variability in cytochrome quences. Pharmacogenomics 2004;5:243^ 72. 18. Evans WE, Relling MV. Moving towards individual- P450 3A5 and in vivo cytochrome P450 3A activity: 4. Kuehl P, Zhang J, Lin Y, et al. Sequence diversity in ized medicine with pharmacogenomics. Nature 2004; some answers but still questions. Clin PharmacolTher CYP3A promoters and characterization of the genetic 429:464 ^ 8. 2004;76:99^103. basis of polymorphic CYP3A5 expression. Nat Genet 19. DaiD,TangJ,RoseR,etal.Identificationofvariants 32. Takano M, Hasegawa R, Fukuda T,Yumoto R, Nagai 2001;27:383^91. of CYP3A4 and characterization of their abilities to J, MurakamiT.Interactionwith P-glycoproteinandtrans- 5. Hustert E, Haberl M, Burk O, et al. The genetic deter- metabolize testosterone and chlorpyrifos. J Pharmacol port of erythromycin, midazolam and ketoconazole in minants of the CYP3A5 polymorphism. Pharmacoge- Exp Ther 2001;299:825^ 31. Caco-2 cells. EurJPharmacol1998;358:289^94. 33. Tolle-Sander S, Rautio J, Wring S, Polli JW, Polli JE. netics 2001;11:773 ^ 9. 20. Eiselt R, Domanski TL, Zibat A, et al. Identification Midazolam exhibits characteristics of a highly permeable 6. LinYS, Dowling AL, Quigley SD, et al. Co-regulation and functional characterization of eight CYP3A4 pro- P-glycoprotein substrate. Pharm Res2003;20:757^64. of CYP3A4 and CYP3A5 and contribution to hepatic tein variants. Pharmacogenetics 2001;11:447 ^ 58. and intestinal midazolam metabolism. Mol Pharmacol 34. Hunter J, Jepson MA,TsuruoT, Simmons NL, Hirst 21. Lee SJ, Bell DA, Coulter S, Ghanayem B, GoldsteinJA. BH. Functional expression of P-glycoprotein in apical 2002;62:162^ 72. Recombinant CYP3A4*17is defective in metabolizing 7. deWildt SN, Kearns GL, LeederJS, van den AnkerJN. membranes of human intestinal Caco-2 cells. Kinetics the hypertensive drug nifedipine and the CYP3A4*17 of vinblastine secretion and interaction with modula- Cytochrome P450 3A: ontogeny and drug disposi- allele mayoccuronthe same chromosome asCYP3A5*3 tion. Clin Pharmacokinet 1999;37:485 ^ 505. tors. J Biol Chem1993;268:14991 ^ 7. representinganew putative defective CYP3Ahaplotype. 35. MarzoliniC,PausE,BuclinT,KimRB.Polymorphisms 8. Evans WE, McLeod HL. Pharmacogenomics: drug JPharmacol ExpTher 2005;313:301^9. disposition, drug targets, and side effects. N Engl J in human MDR1 (P-glycoprotein): recent advances and 22. Ball SE, Scatina J, Kao J, et al. Population distribu- clinical relevance. Clin PharmacolTher 2004;75:13 ^ 33. Med 2003;348:538^49. tion and effects on drug metabolism of a genetic 9. Bates SE, Bakke S, Kang M, et al. A phase I/IIstudy of 36. Hoffmeyer S, Burk O, von Richter O, et al. Functional variant in the 5V promoter region of CYP3A4. Clin polymorphisms of the human multidrug-resistance infusional vinblastine with the P-glycoprotein antago- Pharmacol Ther 1999;66:288 ^ 94. nist valspodar (PSC 833) in renal cell carcinoma. Clin gene: multiple sequence variations and correlation of 23. Baker SD, van Schaik RH, Rivory LP, et al. Factors Cancer Res 2004;10:4724 ^ 33. one allele with P-glycoprotein expression and activity affecting cytochrome P-450 3A activity in cancer 10. Mathijssen RH, de Jong FA, van Schaik RH, et al. in vivo. Proc Natl Acad Sci U S A 2000;97:3473 ^ 8. patients. Clin Cancer Res 2004;10:8341 ^ 50. Prediction of irinotecan pharmacokinetics by use of 37. Sparreboom A, Danesi R, Ando Y, Chan J, Figg WD. cytochrome P450 3A4 phenotyping probes. J Natl 24. Eap CB, BuclinT, Hustert E, et al. Pharmacokinetics Pharmacogenomics of ABC transporters and its role Cancer Inst 2004;96:1585 ^ 92. of midazolam in CYP3A4- and CYP3A5-genotyped in cancer chemotherapy. Drug Resist Updat 2003;6: 11. Lepper ER, Hicks JK, Verweij J, Zhai S, Figg WD, subjects. EurJ Clin Pharmacol 2004;60:231 ^ 6. 71 ^ 84. Sparreboom A. Determination of midazolam in 25. Floyd MD, Gervasini G, Masica AL, et al. Genotype- 38. Goto M, Masuda S, Saito H, et al. C3435T polymor- human plasma by liquid chromatography with phenotype associations for common CYP3A4 and phism in the MDR1gene affects the enterocyte expres- mass-spectrometric detection. J Chromatogr B Ana- CYP3A5 variants in the basal and induced metabolism sion level of CYP3A4 rather than Pgp in recipients of lyt Technol Biomed Life Sci 2004;806:305 ^ 10. of midazolam in European- and African-American men living-donor liver transplantation. Pharmacogenetics 12. Kim JS, Nafziger AN,Tsunoda SM, et al. Limited and women. Pharmacogenetics 2003;13:595 ^ 606. 2002;12:451 ^ 7. samplingstrategy topredictAUCof theCYP3Apheno- 26. Wandel C,Witte JS, Hall JM, Stein CM,Wood AJ, 39. Eap CB, FellayJ, Buclin T,et al. CYP3A activity mea- typingprobemidazolaminadults:applicationtovarious Wilkinson GR. CYP3A activity in African American and sured by the midazolam test is not related to 3435 C>T assay techniques. JClin Pharmacol 2002;42:376 ^ 82. European American men: population differences and polymorphism in the multiple drug resistance trans- 13. van Schaik RH, deWildt SN, Brosens R, van Fessem functional effect of the CYP3A4*1B5V-promoter region porter gene. Pharmacogenetics 2004;14:255 ^ 60. M, van den Anker JN, Lindemans J. The CYP3A4*3 polymorphism. Clin Pharmacol Ther 2000;68:82 ^ 91. 40. Kim RB, Wandel C, Leake B, et al. Interrelationship allele: is it really rare? Clin Chem 2001;47:1104 ^ 6. 27. Lee SJ, Usmani KA, Chanas B, et al. Genetic find- between substrates and inhibitors of human CYP3A 14. van Schaik RH, van der Heiden IP, van den Anker JN, ings and functional studies of human CYP3A5 single and P-glycoprotein. Pharm Res 1999;16:408 ^ 14. Lindemans J. CYP3A5 variant allele frequencies in nucleotide polymorphisms in different ethnic groups. 41. Slaviero KA, Clarke SJ, Rivory LP. Inflammatory Dutch Caucasians. Clin Chem 2002;48:1668 ^ 71. Pharmacogenetics 2003;13:461 ^ 72. response: an unrecognised source of variability in the 15. Goh BC, Lee SC,Wang LZ, et al. Explaininginterindivid- 28. Shih PS, Huang JD. Pharmacokinetics of midazolam pharmacokinetics and pharmacodynamics of cancer ual variability of docetaxel pharmacokinetics and phar- and 1V-hydroxymidazolam in Chinese with different chemotherapy. Lancet Oncol 2003;4:224 ^32. macodynamics in Asians through phenotyping and CYP3A5 genotypes. Drug Metab Dispos 2002;30: 42. Lee HS, Goh BC, Fan L, et al. Phenotyping CYP3A genotyping strategies. JClin Oncol 2002;20:3683^90. 14 91 ^ 6. using midazolam in cancer and noncancer Asian 16. Wong M, Balleine RL, Collins M, Liddle C, Clarke CL, 29. Yu KS, Cho JY, Jang IJ, et al. Effect of the CYP3A5 patients. BrJ Clin Pharmacol 2003;55:270 ^ 7.

Clin Cancer Res 2005;11(20) October 15, 2005 7404 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 24, 2021. © 2005 American Association for Cancer Research. Effect of Common CYP3A4 and CYP3A5 Variants on the Pharmacokinetics of the Cytochrome P450 3A Phenotyping Probe Midazolam in Cancer Patients

Erin R. Lepper, Sharyn D. Baker, Matt Permenter, et al.

Clin Cancer Res 2005;11:7398-7404.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/11/20/7398

Cited articles This article cites 42 articles, 10 of which you can access for free at: http://clincancerres.aacrjournals.org/content/11/20/7398.full#ref-list-1

Citing articles This article has been cited by 7 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/11/20/7398.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/11/20/7398. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.

Downloaded from clincancerres.aacrjournals.org on September 24, 2021. © 2005 American Association for Cancer Research.