Association analysis of polymorphisms in genes related to sunitinib pharmacokinetics M.H.M. Diekstra

Association analysis of polymorphisms in genes related to sunitinib pharmacokinetics

M.H.M. Diekstra

, H.J. Klümpen

, M.P.J.K. Lolkema

, H. Yu

, J.S.L. Kloth

, H. Gelderblom

, R.H.N. van Schaik

, H. Gurney

, J.J. Swen

, A.D.R. Huitema

, N. Steeghs

, R.H.J. Mathijssen

Leiden University Medical Center, Department of Clinical Pharmacy and Toxicology, Leiden, Netherlands

; Academic Medical Center, Amsterdam, Netherlands

; Department of Medical Oncology, University Medical Center Utrecht, Utrecht, Netherlands

; Slotervaart Hospital, Department of Pharmacy & Pharmacology, Amsterdam, Netherlands

; Erasmus MC-Daniel den Hoed Cancer Center, Rotterdam, Netherlands

; Department of Clinical Oncology, Leiden University Medical Center, Leiden, Netherlands

; Westmead Hospital, University of Sydney, Sydney, Australia

; Division of Medical Oncology and Clinical Pharmacology, The Netherlands Cancer Institute-Antoni van Leeuwenhoek Hospital,

Amsterdam, Netherlands

Introduction

 Sunitinib is approved as systemic therapy for metastatic renal cell carcinoma (mRCC), gastrointestinal stromal tumors (GIST) and neuroendocrine tumors of the pancreas (pNET).

 Interpatient variability in pharmacokinetics (PK) of sunitinib and SU12662 is high. Variability in exposure may contribute to the wide range of efficacy and toxicity observed among individual cancer patients.

 Recent studies identified several single nucleotide polymorphisms (SNPs) in candidate genes involved in sunitinib metabolism associated with efficacy and toxicity of sunitinib in mRCC patients.

 These SNPs have not been related to sunitinib and SU12662 PK data.

Conclusions

 SNPs or haplotypes in CYP1A1, CYP3A4, CYP3A5, ABCB1, ABCG2, NR1I2, NR1I3 and POR are not clearly associated with sunitinib or SU12662 clearance.

 Environmental factors, physiological factors and drug-drug interactions also have an effect on CYP3A4 expression and activity and effectuate variability in drug clearances.

 The recently identified CYP3A4*22 SNP potentially has an impact on sunitinib exposure.

Patients and methods

 This multicenter study included 114 oncologic patients treated with sunitinib in seven medical centers integrated in five studies on pharmacokinetics of sunitinib.

 Genotyping was performed using the LightCycler480 RT-PCR (LC480) on 14 selected SNPs related to sunitinib PK (table 3).

 Plasma concentration-time data and pharmacogenetic (PG) data were analysed using the non-linear mixed effect modelling software NONMEM.

 A PK model with a two-compartment model for sunitinib and a one-compartment model for SU12662 was built.

 Clearance of either the parent drug sunitinib or the active metabolite SU12662 (CL,p or CL,m) was used as the PK parameter tested in association analysis.

 Genotypes and haplotypes were separately tested as covariates on clearance for sunitinib and SU12662 in a stepwise manner (forward inclusion at p < 0.05 and backward elimination at p < 0.0005).

Results

Discussion

 This is the first study investigating associations between SNPs in candidate PK genes and clearance of sunitinib and SU12662.

 Reported associations between SNPs in candidate PK genes and sunitinib related toxicities, PFS and OS [1,2,3] are not in line with our findings on SNPs in relation to clearance.

 Previous exploratory studies applied a mild statistical threshold of p < 0.05. To correct for multiple testing, we applied a threshold of p < 0.0005.

 Presence of the CYP3A4*22 allele was associated with a 22.5% decrease in clearance of sunitinib. The CYP3A4*22 allele is also associated with decreased clearance of other CYP3A4 substrates i.e. tacrolimus, midazolam and erythromycin [5,6].

 CYP3A4*22 could potentially be used to individualize sunitinib treatment.

Replication studies in larger groups of patients are needed to verify the role of CYP3A4*22 for sunitinib clearance

M.H.M.Diekstra@lumc.nl Leiden University Medical Center Clinical Pharmacy & Toxicology

0bjective

 To evaluate whether polymorphisms in candidate genes involved in sunitinib metabolism are related to the PK of sunitinib and its active metabolite SU12662.

 6 out of 37 tested genotypes showed a potential association (p<0.05) with clearance of sunitinib or SU12662 in the inclusion step. None of the haplotypes proved significance.

 In the elimination step none of the 6 SNPs reached the significance level of p<0.0005

 The CYP3A4*22 CC genotype was the last to be eliminated from the model (p<0.01)

 A 22.5% decrease in sunitinib clearance was observed in the presence of the CYP3A4*22 T allele (figure 2).

Table 1: Demographics. Abbreviations: ECOG = Eastern Cooperative Oncology Group, (m)RCC = metastatic or non-metastatic renal cell carcinoma, (p)NET = neuroendocrinic tumor located in the pancreas or elsewhere, GIST = gastrointestinal stromal tumor

Figure 2: Empirical Bayes estimates of CL,p and CL,m for base model stratified by different genotypes of CYP3A4*22 intron 6C/T Figure 1: need for SNP-PK association analysis

gene rs number gene rs number

CYP1A1 rs1048943 ABCG2 rs55930652 CYP3A4*22 rs35599367 ABCG2 rs2622604 CYP3A5*3 rs776746 NR1I2 rs3814055 ABCB1 rs1128503 NR1I3 rs2307424 ABCB1 rs2032582 NR1I3 rs2307418 ABCB1 rs1045642 NR1I3 rs4073054 ABCG2 rs2231142 POR*28 rs1057868 Table 3: selected SNPs in candidate PK genes

Sunitinib SU12662

Demographic noted in n(%) or n(range) Gender

Male

Female

75 (65.8%) 39 (34.2%) Age (years) 58.7 (range 27-81) Ethnicity

Caucasian

Asian

Syrian

Hispanic

Unknown (NIB study) 88 (77.2%) 3 (2.6%) 1 (0.9%) 1 (0.9%) 21 (18.4%) Tumor type

(m)RCC

(p)NET

GIST

other solid tumor type 69 (60.5%) 14 (12.3%) 8 (7.0%) 23 (20.2%) ECOG Performance status

0

1

2

Unknown (NIB study) 32 (28.1%) 59 (51.8%) 2 (1.8%) 21 (18.4%) Sunitinib dose (mg)

25 37.5 50 62.5

5 (4.4%) 60 (52.6%) 48 (42.1%) 1 (0.9%)

Table 2: Inclusion step for the correlation between single nucleotide polymorphisms or haplotypes and clearance of the parent drug sunitinib (CL,p) or clearance of the metabolite SU12662 (CL,m). Values in bold are included for analysis, values in bold red are the last to be excluded from the model.

Covariate factors ∆OFV

(CL,p) p-value (CL,p)

∆OFV (CL,m)

p-value (CL,m)

SNPs rs number genotypes

CYP1A1 rs1048943 AA

AG GG

0.047 0.033 -

0.8283 0.8558 -

0.926 0.041 -

0.3359 0.8395 - CYP3A4*22 rs35599367 CC

CT TT

7.451 5.113 -

0.0063 0.0237 -

0.657 0.312 -

0.4176 0.5764 - CYP3A5*3 rs776746 AA

AG GG

- 3.834 6.279

- 0.0502 0.0122

- 4.433 3.347

- 0.0352 0.0673 ABCB1 rs1128503 TT

TC CC

0.505 0.882 3.843

0.4773 0.3476 0.0499

0.562 0.010 0.943

0.4534 0.9203 0.3315 ABCB1 rs2032582 GG

GT TT

0.640 1.435 5.153

0.4237 0.2309 0.0232

0.534 0.558 1.659

0.4649 0.4550 0.1977 ABCB1 rs1045642 TT

TC CC

2.589 0.155 1.079

0.1076 0.6938 0.2989

0.141 0.064 0.005

0.7072 0.8002 0.9436 ABCG2 rs2231142 AA

AC CC

- 2.202 0.836

- 0.1378 0.3605

- 0.013 1.159

- 0.9092 0.2816 ABCG2 rs55930652 CC

CT TT

0.113 1.163 0.995

0.7367 0.2808 0.3185

0.310 0.050 0.986

0.5776 0.8230 0.3207 ABCG2 rs2622604 CC

CT TT

0.579 0.042 1.980

0.4467 0.8376 0.1593

1.914 0.747 3.087

0.1665 0.3874 0.0789 NR1I2 rs3814055 CC

CT TT

0.170 1.745 1.589

0.6801 0.1865 0.2074

0.086 0.293 1.255

0.7693 0.5883 0.2625 NR1I3 rs2307424 TT

TC CC

0.401 1.079 0.828

0.5265 0.2989 0.3628

0.327 0.164 3.370

0.5674 0.6855 0.0663 NR1I3 rs2307418 CC

CA AA

0.613 1.302 -

0.4336 0.2538 -

0.335 0.041 -

0.5627 0.8395 - NR1I3 rs4073054 TT

TG GG

0.160 0.463 0.001

0.6891 0.4962 0.9747

2.159 0.135 0.709

0.1417 0.7133 0.3997 POR*28 rs1057868 CC

CT TT

0.025 0.794 2.700

0.8743 0.3728 0.1003

0.343 0 0.511

0.5581 1 0.4747

Haplotypes

NR1I3

1 or 2 CAT versus no CAT

1.614

0.2039 0.915 0.3387

NR1I3

1 or 2 CAG versus no CAG

0.48 0.4884 0.903 0.3419

ABCB1

1 or 2 TTT versus no TTT

0.204 0.6515 0.017 0.8962

ABCG2 1 or 2 TT versus no TT

0.137 0.7112 2.927 0.0871

References

1. van Erp NP et al. J Clin Oncol 27:4406-4412, 2009, 2. van der Veldt AA et al. Clin Cancer Res 17:620-629, 2011, 3. Garcia-Donas J et al. Lancet Oncol 12:1143-1150, 2011 4. Houk BE et al. Cancer Chemother Pharmacol 66:357-371, 2010 5. Elens et al.

Pharmacogenomics 14(2): 137-149, 2013, 6. Wang D et al. Pharmacogenomics Journal 11:274-286, 2011.

Acknowledgements

We thank dr. T. van der Straaten and R. Baak-Pablo from LUMC for assistance with genotyping and data management.