European Journal of Cardio-Thoracic Surgery Advance Access published February 19, 2013

Mechanical heart valve recipients: anticoagulation in patients with genetic variations of phenprocoumon metabolism† Kerstin Brehma,*, Jenny Schacka, Claudia Heilmanna, Philipp Blankea, Hans Joachim Geisslerb and Friedhelm Beyersdorfa a b

Department of Cardiovascular Surgery, University Heart Center Freiburg—Bad Krozingen, Freiburg, Germany Department of Cardiovascular Surgery, Helios Klinikum Siegburg, Siegburg, Germany

* Corresponding author. Department of Cardiovascular Surgery, University Heart Center Freiburg-Bad Krozingen, Hugstetter Str. 55, 79106 Freiburg, Germany. Tel: +49-761-27024010; fax: +49-761-27025500; e-mail: [email protected] (K. Brehm). Received 14 September 2012; received in revised form 1 December 2012; accepted 10 December 2012

Abstract OBJECTIVES: Oral anticoagulation in mechanical heart valve recipients remains crucial, and vitamin K antagonists (VKA) are still the gold standard. Polymorphisms of vitamin K epoxide oxidase reductase (VKORC) and cytochrome p450 (CYP2C9) were recently found to influence VKA metabolism. We retrospectively investigated the prevalence of these genotypes and associated anticoagulation-related complications in our patients. METHODS: Between August 1998 and August 2008, 563 patients received mechanical heart valve replacement in our institution. Of these, 179 completed a questionnaire on anticoagulation-related complications and consented to genetic analysis. We analysed polymorphisms of VKORC (−1639 G>A; 1173 C>T) and of CYP2C9 (*2, 144 C>T; *3, 359 A>C) by PCR and restriction analysis. RESULTS: For VKORC−1639/1173 alleles, there were 62 (35%) patients with the combination GG/CC, 91 (51%) with GA/CT, 25 (14%) with AA/TT and 1 (1%) with AA/CT. Phenprocoumon (PC) dosage was related to VKORC polymorphism (P < 0.001) with lower doses required for AA/TT patients. Overall, there were 27 severe bleedings and 11 thromboembolic events. For GG/CC, the incidence of major bleeding events and thromboembolic events was 13 and 6%, respectively, and for AA/TT, it was 27 and 12%, respectively. Variation in international normalized ratio (INR) >1.5 was associated with severe bleeding complications (P = 0.025) and GA/CT patients were predisposed to INR variations >1.5 (P = 0.028). No influence of CYP2C9 polymorphism on PC dosage and anticoagulation-related complications was found. CONCLUSIONS: VKORC polymorphism affects PC dosage and anticoagulation-related complication rates in mechanical heart valve recipients. Genotyping may help to identify patients at particular risk of anticoagulation-related complications. Keywords: CYP2C9 • VKORC1 • Phenprocoumon • Mechanical heart valve

INTRODUCTION Long-term anticoagulation with vitamin K antagonists (VKA) is mandatory in patients with mechanical heart valve prostheses. Anticoagulation is most commonly performed with warfarin or phenprocoumon (PC). Achieving stable levels of anticoagulation by adjusting individual dosage requirements is challenging due to broad intra- and inter-individual differences in pharmacokinetics and -dynamics and its narrow therapeutic range. Besides age, body weight, dietary vitamin K intake and co-medication, genetic polymorphisms may influence VKA pharmacokinetics. Single-nucleotide polymorphisms (SNPs) in the cytochrome P450 family 2, subfamily C, polypeptide 9 (CYP2C9) and vitamin

† Presented at the 26th Annual Meeting of the European Association for Cardio-Thoracic Surgery, Barcelona, Spain, 27–31 October 2012.

K epoxide reductase complex subunit 1 (VKORC1) have been identified [1]. While the influence of such polymorphisms is well studied for warfarin or acenocoumarol in patients with mechanical heart valve prostheses, data are missing for PC therapy, which is most commonly used in continental Europe. In Caucasians, common allele variations of CYP2C9*1 wild type are CYP2C9*2 and CYP2C9*3 with frequencies of 10 and 8%, respectively, resulting in lower VKA elimination, so that lower VKA dosages are required [2, 3]. Furthermore, numerous alleles are known for VKORC1, the gene expressing vitamin K epoxide reductase, the main target of VKAs. Patients with −1639 G>A and −1173 C>T alleles usually require lower VKA dosages [4, 5]. Thus, we investigate the influence of genetic polymorphisms in CYP2C9 and VKORC1 regarding individual PC dosage, occurrence of bleeding complications and thromboembolic events in patients with mechanical heart valve prostheses.

© The Author 2013. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.

ADULT CARDIAC

ORIGINAL ARTICLE

European Journal of Cardio-Thoracic Surgery (2013) 1–7 doi:10.1093/ejcts/ezt002

2

K. Brehm et al. / European Journal of Cardio-Thoracic Surgery

MATERIALS AND METHODS

VKORC1 and CYP2C9 genotyping were obtained either at our outpatient clinic or by the patients’ primary physician.

Study population This study is approved by the institutional review board. Written informed consent was obtained from all participants. Between August 1998 and August 2008, 563 patients received surgery, including mechanical heart valve replacement at our institution and were retrospectively reviewed for this study. Fifteen patients younger than 18 years and 146 patients with presurgical PC therapy for known persistent atrial fibrillation were excluded from the study. Patient contact was established beginning in January 2009. Of the 402 patients, 92 were deceased and 41 patients could not be located. The remaining 269 patients were asked to participate in the study, of whom 179 provided informed consent. All 179 participants completed a standardized questionnaire including information on age, sex, weight, height, medical history and co-medication. Further, the following clinical events were recorded in the questionnaire to define the study population group. Bleeding complications were recorded according to the site of bleeding, need for hospital admission or ambulant treatment and divided into minor and major subgroups. Minor bleeding complications were defined as bleedings not requiring medical treatment, whereas major bleeding events were characterized by imminent need for medical therapy. Occurrence of arterial thromboembolic events was recorded and classified as apoplexy, transient ischaemic attack and prolonged reversible neurologic deficit (PRIND). Furthermore, occurrence of pulmonary embolism and deep venous thrombosis were recorded. Target international normalized ratio (INR) range was noted. Administered PC dosages within the last 30 days prior to answering the questionnaire were recorded. The questionnaire was either completed as a telephone interview or at our outpatient clinic. Over-anticoagulation was defined as INR >1.5 above target INR within the last 30 days. Venous blood samples for

Genotyping Genotyping was performed at our institution analogously to previous reports by Sullivan-Klose et al. [6] and Obayashi et al. [7] for CYP2C9 testing and VKORC1 −1639 testing, as well as for VKORC1 −1173 testing. Blood for DNA analysis was collected in Sarstedt Monovette tubes (Sarstedt, Nümbrecht, Germany) containing 1.6 mg/ml ethylenediaminetetraacetic acid (EDTA) and stored at −20°C until further processing. DNA was isolated using the QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany) according to the manufacturers’ instructions. Isolated DNA (100– 300 ng) was subjected to PCR according to standard protocols, using Taq polymerase (Sacace, Como, Italy). PCR products were sequenced. Restriction enzymes were obtained from New England Biolabs (Frankfurt am Main, Germany) and used according to the manufacturers’ instructions. Fragments were separated on 1–3% agarose gels (Peqlab, Erlangen, Germany) and visualized by ethidium bromide staining (Table 1).

Statistical analysis Continuous variables are presented as means ± standard deviation (SD), qualitative variables as absolute frequencies and their percentages. All analyses were performed using SPSS 19.0 (SPSS, Inc., Chicago, IL, USA) or Graph Pad Prism (La Jolla, CA, USA). Non-parametric tests were used to compare groups with regard to PC dosages. Univariate logistic regression analysis was employed to analyse the influence of genotypes on the frequency of overanticoagulation and adverse events. All tests were two sided, and a P value of 65 years of 14.1 ± 5.9 mg, respectively (P < 0.0001). Patients measuring ≤172 cm had an oral intake of PC of 14.6 ± 6.4 mg/week and patients >172 cm had an oral intake of PC of 17.1 ± 7.1 mg/week (P = 0.009). Homozygous carriers of the VKORC1 AA/TT allele required a significantly lower dosage (8.7 ± 2.3 mg/week) than heterozygous GA/CT (15.8 ± 6.7 mg/week) or wild-type GG/CC (19.0 ± 6.0 mg/ week) carriers (P < 0.0001) (Fig. 2). No influence of CYP2C9 polymorphism on weekly PC dosage was observed (*1/*1: 15.8 ± 6.2 mg/week; *1/*2: 16.6 ± 6.7 mg/week; *1/*3: 14.2 ± 9.6 mg/week; *2/*2: 12.0 ± 4.5 mg/week, *2/*3: 10.5 mg/week; P = 0.066). In comparison with combined VKORC1 and CYP2C9 wild-type genotypes (*1/*1 and GG/CC) in all SNPs except *1/*2, a reduction of weekly PC dosage was observed. The dosage-reducing effect was highest in AA/TT/*1/*1 and *1/*3, with a reduction in weekly PC dosage of 54 and 57%, respectively (Fig. 3).

Phenprocoumon dosages Overall, duration of PC therapy added up to a total of 1204.8 years, with a mean of 6.7 ± 2.9 years per patients enrolled. Target

Table 2: Patient demographics and surgical details All patients (n = 179) Age (years) 64 ± 11 (33–90) Gender (male/female) 133/46 Height (cm) 172 ± 9 (149–196) Weight (kg) 82 ± 13 (45–124) PC dosage (mg/week) 15.9 ± 6.8 (4.5–54) Surgery performed AVR 104/179 AVR + CABG 27/179 AVR + ascending aortic 14/179 replacement AVR + CABG + ascending aortic replacement 1/179 AVR + MVR 7/179 MVR 23/179 MVR + CABG 3/179

Percentages (%)

Figure 1: Combinations and their frequencies of CYP2C9 and VKORC1−1639 genotype occurrence (n = 175)

59 15 8

1 4 13 2

AVR: aortic valve replacement; CABG: coronary artery bypass grafting; MVR: mitral valve replacement; PC: Phenprocoumon. Figure 2: VKORC1 −1639 genotypes and PC dosage

ADULT CARDIAC

K. Brehm et al. / European Journal of Cardio-Thoracic Surgery

4

K. Brehm et al. / European Journal of Cardio-Thoracic Surgery

Bleeding events A total of 60 bleeding events were reported by 54 patients. There were 27 patients with minor bleeding events and 27 with major bleeding events. Reported bleeding complications were epistaxis (30 events in 24 patients), intramuscular bleedings (n = 9), gastrointestinal bleeding (n = 8), postoperative bleeding after noncardiac procedures (n = 7), cerebral haemorrhage (n = 3), kidney haemorrhages (n = 1), spinal haemorrhages (n = 1) and bleeding after tongue bite (n = 1). The relative rate of bleeding complications was 4% per patient-year with 2% per patient-year for minor and 2% per patient-year for major bleeding complications. No significant correlation between bleeding complications and CYP2C9 and VKORC1 genotypes was observed (Table 3). However, homozygous VKORC1 AA/TT patients showed a higher tendency to major bleeding complications (7 of 26; 27%), than GA/CT patients (12 of 91; 13%) and wild-type GG/CC patients (8 of 62; 13%) (Fig. 4). Severe over-anticoagulation (INR variation > 1.5) was associated with major bleeding complications (P = 0.025, OR 3.6, 95% CI 1.2– 11.2). Patients with the VKORC1 GA/CT genotype had a significantly increased risk of severe over-anticoagulation (P = 0.028, OR 5.4, 95% CI 1.2–24.1).

Thromboembolic events The overall incidence of thromboembolic events under PC therapy was 6% (11 of 179 patients), relating to 1% per patientyear. Seven patients suffered from ischaemic apoplexy (7 of 179, 4%) and 4 had a transient ischaemic episode (4 of 179; 2%). No significant differences were detected in the occurrence of thromboembolic events for CYP2C9 and VKORC1 genotypes (Table 4). However, CYP2C9 *1/*2 and VKORC1 AA/TT patients had higher tendency to thromboembolic events (9 and 11% respectively).

DISCUSSION We investigated the influence of genetic polymorphisms in CYP2C9 and VKORC1 on individual PC dosages, occurrence of

bleeding complications and thromboembolic events in 179 patients with mechanical heart valve prostheses. The allele frequencies observed in our study population for CYP2C9 and VKORC1 were in accordance with other studies on Caucasians [8, 9]. Most frequently, a combination of wild-type *1/*1 CYP2C9 and heterozygous GA mutation in VKORC1 was observed. VKORC1 polymorphisms significantly affected the PC dosages needed to reach stable INR values, whereas no significant effect was found for CYP2C9 polymorphisms. This is in line with the findings by Geisen et al. [10], who found that 38% of inter-individual variability in PC dosages is solely attributable to the VKORC 1 polymorphism. Furthermore, we observed significantly lower weekly PC doses in VKORC1 polymorphism carriers (AA/TT and GA/CT), in line with findings by others [4, 11]. In contrast, we did not observe a significant influence of CYP2C9 genotypes on PC dosages. This is in line with similar findings in various patient populations [3, 12]. Interestingly, significant effects of CYP2C9 polymorphism on the dosages needed to achieve a stable anticoagulation have been observed for other VKAs, in particular, for warfarin [13]. This was not observed in our study, most likely due to differences in drug elimination. While 40% of unmetabolized PC is eliminated via renal secretion, warfarin is almost completely metabolized. In addition, CYP2C9-catalysed biotransformation plays a lessimportant role in the elimination of PC, as CYP3A4 is also involved in its metabolism [14]. Due to the longer elimination half-life of PC compared with warfarin, dosage adjustments in patients treated with PC are associated with a delayed INR response and might predispose to a greater risk of overanticoagulation. However, patients with long-term use of PC are more likely to have INR values within the therapeutic range, compared with patients who receive VKAs with shorter half-lives [15]. For combined genotypes, Schalekamp et al. [11] observed a mutual interference of CYP2C9 and VKORC1 polymorphisms in PC dosage. Patients with VKORC1 wild-type and CYP2C9 polymorphism required up to 30% less PC than CYP2C9*1 patients. Additional SNPs in VKORC1 resulted in less difference in dosage. However, this was not observed in the current investigation.

Figure 3: Combined CYP2C9 and VKORC1 −1639 genotypes and average weekly PC dosage

5

K. Brehm et al. / European Journal of Cardio-Thoracic Surgery

Odds ratio Minor bleeding events CYP2C9 Wild-type *1/*1 SNP*2 SNP*3 VKORC1 Wild-type GG GA AA Major bleeding events CYP2C9 Wild-type *1/*1 SNP*2 SNP*3 VKORC1 Wild-type GG GA AA

16/109 (15%) 5/42 (12%) 5/24 (21%) 12/62 (19%) 13/91 (14%) 2/26 (8%)

1.27 0.65

1.44 2.88

P

0.8 0.5

0.5 0.2

Table 4: CYP2C9 and VKORC1−1639 genotypes and thromboembolic events

CI

0.4–3.7 0.2–2.0

0.6–3.4 0.6–14

CYP2C9 Wild-type *1/*1 SNP*2 SNP*3 VKORC1 Wild-type GG GA AA

6/109 (5%) 4/42 (9%) 1/24 (4%) 4/62 (6%) 4/91 (4%) 3/26 (11%)

Odds ratio

P

CI

0.55 1.34

0.5 1.0

0.1–2.1 0.1–12

1.5 0.53

0.7 0.4

0.4–6.2 0.1–2.5

CI: confidence intervals. 17/109 (16%) 7/42 (17%) 3/24 (12%)

0.92 1.29

1.0 1.0

0.3–2.4 0.3–4.8

8/62 (13%) 12/91 (13%) 7/26 (27%)

0.98 0.4,

1.0 0.1

0.4–2.5 0.1–1.3

CI: confidence intervals. Total number of bleeding events in CYP2C9 patients is 53, instead of 54, as in 1 patient, no definite CYP2C9 genotype was found.

Figure 4: VKORC1 −1639 genotypes and major bleeding events

For clinical routine, genotype-guided algorithms could be used to reduce the time needed to determine individual daily VKA dosage and might help to avoid over-anticoagulation. Algorithms concerning warfarin dosage and CYP2C9 and VKORC1 genotyping have been developed and tend to have a positive effect on predicting warfarin dosage [9, 16]. However, dosage-prediction models for warfarin are not applicable to PC. A dosage algorithm for different CYP2C9 and VKORC1 genotypes has been established by van Schie et al. [17] for the initiation phase of PC therapy (3 months). Further investigations on genotype-based dosage vs non-genotype dosage are in progress. As CYP2C9 SNPs did not have a significant influence on PC dosage, VKORC1 genotyping alone might be sufficient. In the current investigation, bleeding complications were not significantly associated with CYP2C9 polymorphisms. This is in contrast to observations by Hummers-Pradier et al. who found that CYP2C9*3 was associated with a triple increase in bleeding

risk [12]. Visser et al. [18] found an increased bleeding risk for CYP2C9 SNPs and acenocoumarol but not for PC. Regarding VKORC1 SNPs, relevant effects on bleeding complications depend on the applied VKA therapy. PC therapy is associated with a higher risk of bleeding in patients with at least one mutation in the VKORC1 gene [4]. In contrast, no association of warfarin or acenocoumarol therapy in VKORC1 SNP patients with the occurrence of bleeding complications has been found [4, 19]. Although we did not observe any significant association of VKORC1 polymorphism and occurrence of bleeding complication, we noticed a higher incidence of major bleeding complications in VKORC1 AA/TT patients compared with heterozygous or wild-type patients. The incidence of thromboembolic events was 6%. Although no significant association between CYP2C9 and VKORC1 genotypes was observed, CYP2C9 *1/*2 and VKORC1 AA/TT patients had a higher incidence of thromboembolic events (9 and 11%, respectively). This is in line with findings by Giansante et al. [20], who found an increased risk of thromboembolism in patients carrying the A allele while taking warfarin. However, a study investigating thromboembolic complications with PC is lacking. Importantly, variations in INR >1.5 were associated with severe bleeding complications. Interestingly, GA/CT patients were predisposed for INR variations >1.5. Therefore, one might state that GA/CT mutations in VKORC1 are at higher risk of INR variations >1.5 and consecutively at higher risk for severe bleeding complications. Although Schalekamp et al. observed an increased risk of over-anticoagulation in carriers of the CYP2C9*2 or *3 allele during the initiation phase of PC therapy [21], no influence of CYP2C9 genotypes on over-anticoagulation was observed in our study. This may be due to the differences in the observed time point of over-anticoagulation, as it was in the maintenance phase in our study. This is consistent with Luxembourg et al. [22], where CYP2C9 defect carriers were not associated with adverse outcome parameters in the initiation and maintenance phase, while VKORC1 AA carriers showed a prolonged initiation phase and a high risk of over-anticoagulation. In contrast to acenocoumarol and warfarin, which have been thoroughly studied in mechanical heart valve recipients in regard to CYP2C9 and VKORC1 genotyping, data on PC are missing. However, there seems to be a difference between anticoagulation and genotype influence of acenocoumarol and

ADULT CARDIAC

Table 3: CYP2C9 and VKORC1−1639 genotypes and bleeding events

6

K. Brehm et al. / European Journal of Cardio-Thoracic Surgery

warfarin on one hand and PC on the other. CYP2C9 genotypes seem not to have much impact on complication and effective dosage under PC therapy. However, VKORC1 genotyping seems reasonable as PC dosage in mutations is much lower and INR variations >1.5 are much more frequent in GA/CT patients. VKORC1 gene polymorphism warrants adjustment of PC dose to optimize its efficacy and minimize bleeding. VKORC1 genotyping before anticoagulant therapy might help to guide PC dosage in patients with mechanical heart valves, but further trials for dosage algorithms are needed.

Study limitations Limitations are associated with this study. It is a retrospective study with a potential bias concerning the information on adverse events given by the patient in the questionnaire. However, rates of adverse events do not differ from those previously described for mechanical heart valve recipients. Prospective clinical studies are required to assess dosing algorithms and adverse events.

CONCLUSIONS In conclusion, VKORC polymorphism affects PC dosage requirements and anticoagulation-related complication rates in mechanical heart valve recipients. Genotyping may help to identify patients at particular risk of anticoagulation-related complications.

Funding This study was supported by Edwards Lifesciences LLC. Conflict of interest: none declared.

REFERENCES [1] Oldenburg J, Bevans CG, Fregin A, Geisen C, Muller-Reible C, Watzka M. Current pharmacogenetic developments in oral anticoagulation therapy: the influence of variant VKORC1 and CYP2C9 alleles. Thromb Haemost 2007;98:570–8. [2] Higashi MK, Veenstra DL, Kondo LM, Wittkowsky AK, Srinouanprachanh SL, Farin FM et al. Association between CYP2C9 genetic variants and anticoagulation-related outcomes during warfarin therapy. JAMA 2002; 287:1690–8. [3] Visser LE, van Vliet M, van Schaik RH, Kasbergen AA, De Smet PA, Vulto AG et al. The risk of overanticoagulation in patients with cytochrome P450 CYP2C9*2 or CYP2C9*3 alleles on acenocoumarol or phenprocoumon. Pharmacogenetics 2004;14:27–33. [4] Reitsma PH, van der Heijden JF, Groot AP, Rosendaal FR, Buller HR. A C1173T dimorphism in the VKORC1 gene determines coumarin sensitivity and bleeding risk. PLoS Med 2005;2:e312. [5] Rieder MJ, Reiner AP, Gage BF, Nickerson DA, Eby CS, McLeod HL et al. Effect of VKORC1 haplotypes on transcriptional regulation and warfarin dose. N Engl J Med 2005;352:2285–93. [6] Sullivan-Klose TH, Ghanayem BI, Bell DA, Zhang ZY, Kaminsky LS, Shenfield GM et al. The role of the CYP2C9-Leu359 allelic variant in the tolbutamide polymorphism. Pharmacogenetics 1996;6:341–9. [7] Obayashi K, Nakamura K, Kawana J, Ogata H, Hanada K, Kurabayashi M et al. VKORC1 gene variations are the major contributors of variation in warfarin dose in Japanese patients. Clin Pharmacol Ther 2006;80:169–78. [8] Lee CR, Goldstein JA, Pieper JA. Cytochrome P450 2C9 polymorphisms: a comprehensive review of the in-vitro and human data. Pharmacogenetics 2002;12:251–63.

[9] Wadelius M, Chen LY, Lindh JD, Eriksson N, Ghori MJ, Bumpstead S et al. The largest prospective warfarin-treated cohort supports genetic forecasting. Blood 2009;113:784–92. [10] Geisen C, Luxembourg B, Watzka M, Toennes SW, Sittinger K, Marinova M et al. Prediction of phenprocoumon maintenance dose and phenprocoumon plasma concentration by genetic and non-genetic parameters. Eur J Clin Pharmacol 2011;67:371–81. [11] Schalekamp T, Brasse BP, Roijers JF, van Meegen E, van der Meer FJ, van Wijk EM et al. VKORC1 and CYP2C9 genotypes and phenprocoumon anticoagulation status: interaction between both genotypes affects dose requirement. Clin Pharmacol Ther 2007;81:185–93. [12] Hummers-Pradier E, Hess S, Adham IM, Papke T, Pieske B, Kochen MM. Determination of bleeding risk using genetic markers in patients taking phenprocoumon. Eur J Clin Pharmacol 2003;59:213–9. [13] Gage BF. Pharmacogenetics-based coumarin therapy. Hematology Am Soc Hematol Educ Program 2006;467–73. [14] Ufer M, Svensson JO, Krausz KW, Gelboin HV, Rane A, Tybring G. Identification of cytochromes P450 2C9 and 3A4 as the major catalysts of phenprocoumon hydroxylation in vitro. Eur J Clin Pharmacol 2004;60: 173–82. [15] Rombouts EK, Rosendaal FR, van der Meer FJ. Subtherapeutic oral anticoagulant therapy: frequency and risk factors. Thromb Haemost 2009; 101:552–6. [16] Anderson JL, Horne BD, Stevens SM, Woller SC, Samuelson KM, Mansfield JW et al. A randomized and clinical effectiveness trial comparing two pharmacogenetic algorithms and standard care for individualizing warfarin dosing (CoumaGen-II). Circulation 2012;125:1997–2005. [17] van Schie RM, Wessels JA, le Cessie S, de Boer A, Schalekamp T, van der Meer FJ et al. Loading and maintenance dose algorithms for phenprocoumon and acenocoumarol using patient characteristics and pharmacogenetic data. Eur Heart J 2011;32:1909–17. [18] Visser LE, van Schaik RH, van Vliet M, Trienekens PH, De Smet PA, Vulto AG et al. The risk of bleeding complications in patients with cytochrome P450 CYP2C9*2 or CYP2C9*3 alleles on acenocoumarol or phenprocoumon. Thromb Haemost 2004;92:61–6. [19] Limdi NA, Beasley TM, Crowley MR, Goldstein JA, Rieder MJ, Flockhart DA et al. VKORC1 polymorphisms, haplotypes and haplotype groups on warfarin dose among African-Americans and European-Americans. Pharmacogenomics 2008;9:1445–58. [20] Giansante C, Fiotti N, Altamura N, Pitacco P, Consoloni L, Scardi S et al. Oral anticoagulation and VKORC1 polymorphism in patients with a mechanical heart prosthesis: a 6-year follow-up. J Thromb Thrombolysis 2012. [21] Schalekamp T, Oosterhof M, van Meegen E, van Der Meer FJ, Conemans J, Hermans M et al. Effects of cytochrome P450 2C9 polymorphisms on phenprocoumon anticoagulation status. Clin Pharmacol Ther 2004;76:409–17. [22] Luxembourg B, Schneider K, Sittinger K, Toennes SW, Seifried E, Lindhoff-Last E et al. Impact of pharmacokinetic (CYP2C9) and pharmacodynamic (VKORC1, F7, GGCX, CALU, EPHX1) gene variants on the initiation and maintenance phases of phenprocoumon therapy. Thromb Haemost 2011;105:169–80.

APPENDIX. CONFERENCE DISCUSSION Dr K. Hekmat (Cologne, Germany): You presented a study with 179 patients with mechanical heart valves and did a genotyping to identify patients at risk for anticoagulation-related complications. As we all know, age, body weight and height, and vitamin K intake and comedication may influence the pharmacokinetics of vitamin K antagonists. I have two questions. Did you also calculate the odds ratio of these variables, I mean, age, height and so forth? And the second question, why did you include mitral valve patients in your analysis since thromboembolic events may be more frequent in these patients? Dr Brehm: Regarding your first question, we did calculate the odds ratio, and only age and height had an influence. We looked at the comedication as well, and no influence was found. I am aware that other studies have shown differences in factors like comedication, age, height, and weight. The second question was? Dr Hekmat: The mitral valve patients. Dr Brehm: Mitral valve patients are known to have a larger number of thromboembolic events, but we did not see that in our study. The rate of our thromboembolic events was 0.6% per patient year on phenprocoumon therapy, and that is consistent with rates described in the literature. So I do not think that we had a higher number of thromboembolic events by including mitral valve patients.

Dr Hekmat: So when you look at the odds ratio of the age and height and so on, is the odds ratio higher? Is there more impact of age and so on, as for polymorphism? Dr Brehm: No. Polymorphism has much more impact. In particular, the VKORC1 genotype is responsible for about 30 to 38% of the difference in phenprocoumon dosage. Dr J. Grau (Ridgewood, NJ, USA): If I understand correctly, you correlate your genotyping to thromboembolic events and bleeding, right? Dr Brehm: Yes, we did. Dr Grau: Did you do any correlation of your genotyping for those two levels to the INR response in those patients? In other words, basically what you are really looking at is whether these people behaved the same or differently than patients without those genotypical abnormalities.

7

Dr Brehm: Yes. We just had a look at the maintenance phase, not the initiation phase of phenprocoumon therapy, as the data of the initiation phase was not available in our patients. So, strictly speaking, the study was on patients in the maintenance phase of phenprocoumon therapy. Dr M. Jahangiri (London, UK): Do you do elective genotyping now? Dr Brehm: No, not yet. Dr Jahangiri: Is it because it is expensive? Dr Brehm: It is expensive, but routine genotyping is under discussion in the United States. The FDA refers in its prescription information of warfarin to genotyping, as genotyping may be helpful in reducing the risk of warfarin therapy. Personally I think genotyping for warfarin or phenprocoumon in Germany is coming, but we are not doing it at the moment.

ADULT CARDIAC

K. Brehm et al. / European Journal of Cardio-Thoracic Surgery