Bone Marrow Transplantation (2007) 40, 659–664 & 2007 Nature Publishing Group All rights reserved 0268-3369/07 $30.00 www.nature.com/bmt

ORIGINAL ARTICLE 2C19 loss-of-function polymorphism is associated with an increased treatment-related mortality in patients undergoing allogeneic transplantation

AH Elmaagacli, M Koldehoff, NK Steckel, R Trenschel, H Ottinger and DW Beelen

Department of Bone Marrow Transplantation, University Hospital of Essen, Essen, Germany

The polymorphic expression of CYP2C19 causes characterized with a large number of genetic individual variability in drug metabolism and thereby in polymorphisms is cytochrome P450 (CYP) 2C19.1 Pre- pharmacologic and toxicologic responses. We genotyped vious studies on CYP2C19 using probe drugs such as 286 patients and their donors for the CYP2C19 gene who S- showed that individuals could be classified underwent allogeneic transplantation for various diseases into three different groups: poor metabolizers (PMs), and analyzed their outcome. Patients were classified as: intermediate metabolizers (IMs) and extensive metabolizers poor metabolizers (PMs; 3.1%), intermediate metaboli- (EMs). PMs have a genetically determined absence of active zers (IMs; 24.5%) and extensive metabolizers (EMs; enzymes, which is the cause for a slower metabolism of 72.5%). Patients genotyped as PMs had significant higher active drugs and is associated with prolonged side effects.1 hepato- and nephrotoxicities compared to IMs or EMs. If prodrugs are applied, the metabolism in their active form Maximum bilirubin and serum creatinine levels measured is delayed and reduced effectiveness may occur. Among the after transplant were approximately twofold higher than drugs which are metabolized by CYP2C19 enzymes are those of EMs or IMs. The increased toxicity resulted in an cytostatics, proton pump inhibitors, , seda- increased 4-year estimate for transplant-related mortality tive, beta blockers, antiviral and anti-fungal agents.2–9 (TRM) with 50718.6% for PMs compared to Recently, it was demonstrated that PMs of CYP2C19 25.173.7% for EMs (Po0.018) and 22.7 75.6% for showed a decrease in platelet responsiveness to clopido- IMs (Po0.042), whereas no significant influence for grel.10 In another study, it was reported that patients relapse rate, overall survival or incidence of acute graft- treated with antidepressants and genotyped as PMs for versus-host disease grade 2–4 were found between the CYP2D6 and CYP2C19, remained in hospital for longer groups. Multivariate analysis including all potential (median 57.5 versus 40.0 days) compared to IMs.8 factors that might influence TRM confirmed that the Patients undergoing allogeneic transplantation are trea- genotype of CYP2C19 is an independent factor, which ted with more drugs, that is, immunosuppressives, chemo- influenced TRM significantly. These results suggest that therapeutical agents or anti-infectious medication, each of genotyping for CYP450 2C19 can helpto identify patients them pivotal for the outcome of transplant. However, we with higher risk for TRM. hypothesized that altered drug metabolism in patients who Bone Marrow Transplantation (2007) 40, 659–664; underwent allogeneic transplantation might influence doi:10.1038/sj.bmt.1705786; published online 6 August 2007 transplant outcome. In a retrospective single-center study, Keywords: CYP2C19; poor metabolizers; transplant this paper evaluates whether gene polymorphisms of CYP2C19 influence the incidence of acute graft- versus-host disease (GVHD) after transplant, and/or influence treatment-related mortality (TRM) and thereby Introduction the outcome of allogeneic stem cell transplantation.

The polymorphic expression of drug-metabolizing enzymes Patients and methods is one of the major factors, which cause individual variability in drug metabolism and thereby in pharma- Patients cologic and toxicologic responses. One of the best We included a total of 286 patients (and donors) for the CYP2C19 gene polymorphism analysis, who were trans- planted at the University Hospital of Essen between Correspondence: Dr AH Elmaagacli, Department of Bone Marrow October 1998 and December 2004. Genotyping was Transplantation, University Hospital of Essen, Hufelandstr. 55, 45122 performed without knowledge of GVHD status or outcome Essen, Germany. E-mail: [email protected] of the patients analyzed. All aspects of this study were Received 9 March 2007; revised 6 June 2007; accepted 11 June 2007; approved by the Institutional Review Board on Medical published online 6 August 2007 Ethics at the University Hospital of Essen. Loss-of-function allele of CYP2C19 and transplant-associated mortality AH Elmaagacli et al 660 Pre-transplant histocompatibility testing of patients and Analysis and measurement of blood parameters donors was performed at low-resolution human leukocyte Total bilirubin, g-glutamyl , prothrombin time, antigen (HLA)-A, -B, -C and high-resolution HLA-DRB1, activated partial thromboplastin, fibrinogen and creatinine DQB1 level DNA-based typing according to standard were determined at baseline, and daily thereafter, until methods.11 at least day þ 30 after HSCT. Weight and abdominal circumference were measured daily. Conditioning regimen The conditioning regimen consisted of intravenous cyclo- Isolation of genomic DNA phosphamide in combination with fractioned total body DNA was prepared from peripheral blood mononuclear irradiation (TBI) or oral busulfan (BU), in combination cells obtained from the donor and patient before the with or treosulfan in combination with transplant using the phenol/chloroform method.14 cyclophosphamide or fludarabine as published earlier.12 Patients who underwent transplantation of highly enriched CD34 þ cells received a conditioning regimen with frac- Genotyping for CYP2C19 tioned TBI, cyclophosphamide and thiotepa as published Polymorphisms of CYP2C19 were determined by use of the earlier.12 Patients scheduled for peripheral blood stem cell hybridization probe format (LightCycler CYP2C19 Muta- transplantation (PBSCT) with reduced conditioning re- tion Detection Kit with specific primers; Roche Applied ceived BU (1 mg/kg of body weight every 6 h over 2days) in Science; Mannheim, Germany). Following genotypes were combination with fludarabine (30 mg/m2 of body surface determined: CYP2C19*1/*1 (EMs); CYP2C19*1/*2 (IMs), area over 5 days). CYP2C19*1/*3 (IMs), CYP2C19*2/*2 (PMs), CYP2C19*2/ GVHD prophylaxis consisted of methotrexate (MTX) *3 (PMs), CYP2C19*3/*3 (PMs). PCR and subsequent and cyclosporine (CSA) in patients who received an melting curve analysis were performed using the Lightcycler unmanipulated graft.12 In the patients who underwent a device (Roche Applied Science) and software. Control CD34 þ -PBSCT, no further GVHD prophylaxis was given. samples confirmed by sequencing were included in each run. In vivo T-cell depletion was performed using Campath-1 H (10 or 20 mg for 5 days) in combination with CSA or (rabbit) anti T-lymphocyte globuline (ATG-S, Fresenius, Definition of the study groups Bad Homburg, Germany) (10 mg/kg of body weight for Patients with wild type (EMs) of the analyzed gene were 4 days) in combination with CSA and MTX. defined as group 1. Patients with heterozygous Patient demographics are summarized in Table 1. (IMs) of the analyzed gene were designated as group 2, while patients with homozygous mutations (PMs) of the analyzed gene at the donor side only were named group 3. Supportive care Besides these three groups, we also investigated the possible All patients were isolated in reverse isolation rooms influence of donor genotype on patient outcome. equipped with high-efficiency particular filtration systems and received prophylactic metronidazole 400 mg three times daily and oral ciprofloxacin, 750 mg twice daily and Statistics fluconazole, 200 mg once daily. Cumulative estimates (7standard errors) were calculated by the Kaplan–Meier method.15 Differences between time- to-event distribution functions were compared by a log- Clinical study endpoints rank test (Mantel–Haenszel).16 A stepwise proportional Neutrophil engraftment was defined as the first of three hazards general linear model (PHGLM) analysis was used consecutive days with an absolute neutrophil count greater to evaluate interactions of different covariates on the than 0.5 Â 109/l. Acute GVHD was graded according to analytical endpoint of TRM. Covariates in PHGLM standard criteria.13 Chronic GVHD was assessed in patients analyses were stratified according to severe acute GVHD, alive after 100 days. TRM was defined as death with no age ( 40 or 440 years), gender constellation (male relapse. Overall survival (OS) was defined as survival from p patient with female donor or other), PM (mutations at transplantation without death from any cause. patient site) versus others, HLA-constellation between patient and donor, sibling donor versus unrelated donor, Diagnostic criteria for veno-occlusive disease graft-type (PBSC or bone marrow), T-cell depleted graft Veno-occlusive disease (VOD) was suspected clinically versus non T-cell depleted graft, and disease stage according to the modified Seattle criteria which require (advanced versus early disease stages) were analyzed. the presence of at least two of the following three clinical Conditional risk ratios (RR) and their 95% confidence findings before day þ 30 after hematopoietic stem cell intervals (95% CI) were derived from PHGLM analyses transplantation (HSCT): (a) jaundice with bilirubin after adjustment for significant covariates in the model. 434 mmol/l, (b) painful hepatomegaly and (c) fluid Statistical analysis was performed using the SPSS software retention 45% of the body weight at the time of HSCT. version 12.0. In all patients, the diagnosis of VOD was confirmed either Only patients surviving for more than 30 days were by demonstration of ascites and inversion of portal venous included in the analysis of acute GVHD. A minimum flow using Doppler sonography, or by liver biopsy or of 100 days of follow-up was the criterion for chronic autopsy. GVHD.

Bone Marrow Transplantation Loss-of-function allele of CYP2C19 and transplant-associated mortality AH Elmaagacli et al 661 Table 1 CYP2C19: demographic and treatment characteristics of patients (number)

EM-patients IM-patients PM-patients P-value* Group 1 Group 2 Group 3

Number of patients (n) 207 70 9

Median age (range)(years) Patients 40 (17–67) 41 (20–58) 44 (26–59) NS Donors 38 (16–65) 37 (18–65) 38 (27–47) NS

Gender Female/male 108/109 40/30 2/7 NS

Gender of donor and recipient Male with female donor 124 0 NS All others 195 66 9 NS

Diseases Acute leukemia (AML, ALL) 120 34 4 CML 43 17 1 MDS 19 9 2 NHL 11 3 1 MM 5 4 0 SAA 5 1 1 Others 4 20

Disease phases Early phase 93 34 3 NS Advanced phases 114 36 6 NS

Donor HLA-type HLA-identical 171 59 7 NS HLA-mismatched 37 11 2NS

Donor type Sibling 13244 4 Po0.05 (group 3 versus others) Unrelated 75 26 5 NS

Graft source PBSCT 159 53 9 NS BMT 48 17 0 NS

GVHD prophylaxis CSA and MTX 145 48 4 NS T-cell depletion 4218 5 NS CSA alone 20 4 0 NS

Conditioning regimens TBI consisting 144 49 5 NS alone 63 21 4 NS Median follow-up (in months) 65 69 21 Po0.02(group 3 versus others)

Abbreviations: ALL ¼ acute lymphocytic leukemia; AML ¼ acute myelocytic leukemia; BMT ¼ bone marrow transplantation; CML ¼ chronic myelocytic leukemia; CSA ¼ cyclosporine; EM ¼ extensive metabolizer; HLA ¼ human leukocyte antigen; GVHD ¼ graft-versus-host disease; IM ¼ intermediate metabolizer; MDS ¼ myelodysplastic syndrome; MTX ¼ methotrexate; MM ¼ multiple myeloma; NHL ¼ non-Hodgkin’s lymphoma; PBSCT ¼ peripheral blood stem cell transplantation; PM ¼ poor metabolizer; SAA ¼ severe aplastic anemia A; TBI ¼ total body irradiation. *Po0.05.

Results Incidence of GVHD No statistic differences were found in the incidence of Frequency of EMs, IMs and PMs acute GVHD grade 2–4 for EMs, IMs and PMs with 28.2, Of the patients 207 (72.4%) were genotyped as EMs, 70 32.8 and 25%, respectively. The genotype of donors also (24.5%) as IMs and 9 (3.1%) as PMs. Of the donors, 204 had no significant influence on the incidence of GVHD (71.3%) were EMs and had a homozygous wild-type gene, grade 2–4 in patients. Patients with EM donors and those 68 (23.8%) were genotyped as IMs and had a hetero- with IM donors did not differ statistically in the incidence zygous genotype and 14 donors (4.8%) were typed as PMs of acute GVHD grade 2–4 (27.4 % EM versus 22.4% with a homozygous gene . Calculated allele IM). There were also no differences found in the incidence frequencies did not differ much from those reported of acute GVHD grade 2–4 for PM-patients with 25%. earlier by other studies for Caucasians, as shown in However, the highest incidence of acute GVHD grade 2–4 Table 2.17,18 of 35.7% occurred in patients with PM-donors, although

Bone Marrow Transplantation Loss-of-function allele of CYP2C19 and transplant-associated mortality AH Elmaagacli et al 662 Table 2 Comparison of CYP2C19 allele frequency between patients, donors and Caucasian controls

CYP2C19

*1 in % (95% CI) *2 in % (95% CI) *3 in % (95% CI) P-value

Patients (n ¼ 286) 84.8 (83.3–89.5) 12.1 (9.0–15.2) 3.1 (0–4.7) NS Donors (n ¼ 286) 83.8 (82.7–89.3) 11.7 (8.4–15.0) 4.5 (0–4.5) NS Caucasians control group 86.4 (82.9–89.9) 12.7 (9.2–16.2) 0.9 (0–4.4) NS (n ¼ 546)17 Caucasians control group 86.7 (83.7–89.2) 8.8 (6.7–11.4) 5.3 (3.7–7.4) NS (n ¼ 302)18

Abbreviations: CI ¼ confidence interval; n ¼ total number of alleles.

this was not statistically significant when compared to Maximum creatinine in mg/dl (median)

IMs or EMs. P<0.05 The incidence of chronic GVHD (extensive and limited P<0.04 together) was almost identical in the four study groups, 2 1.9 with 73% in group 1, 68% in group 2and 66% in group 3. 1.8 Before transplant ± 1.2 1.6 After transplant Influence of CYP2C19 gene polymorphism on 1.4 hepatotoxicity and nephrotoxicity after transplant 1.2 First, we compared the bilirubin and creatinine levels in 1 1.06 0.99 PMs before transplant and defined them as baseline values. 0.8 ± ± 0.85 0.78 0.41 0.81 0.81 ± ± Thereafter, we evaluated the maximum bilirubin and 0.6 ± 0.20 0.22 creatinine values after transplant for all study groups. We 0.4 0.25 found a more than threefold increase of bilirubin after 0.2 transplant from baseline in PM patients (Po0.005). Serum 0 creatinine also increased more than twofold from baseline EMs IMs PMs after transplant as shown in Figures 1 and 2. We then compared the maximum bilirubin and creatinine values of Figure 1 Maximum creatinine levels in mg/dl before and after transplant for each study group. Poor metabolizers had the highest bilirubin and PMs with those of IMs and EMs. As expected, the highest creatinine values after transplant compared to extensive metabolizers or maximum bilirubin and maximum creatinine levels after intermediate metabolizers. transplant were documented in PMs as shown in Figures 1 and 2. The maximum bilirubin (median) was more than Maximum bilirubin in mg/dl (median) twofold higher than that of EMs and IMs (Po0.05). PMs had also the highest increase in serum creatinine levels P<0.04 compared to IMs and EMs. 2.5 VOD defined as mentioned above occurred slightly more P<0.02 often in PM-patients compared to the other study groups. Before transplant 2 VOD was diagnosed in 2 of 9 PM-patients (22%) and 5 of 38 After transplant 2.0 ± IM-patients evaluated (13%) and 13 of 105 EM patients 20.1 evaluated (12.4%), but these differences were not significant. 1.5

1 1.01 Treatment-related mortality, overall survival and ± 0.81 occurrence of relapse 8.1 ± 0.5 0.55 0.62 0.53 1.52 ± Four-year estimate for TRM was highest in PM-patients ± ± 0.09 with 50718.6% compared to EM patients (24.373.6% 0.25 0.21 0 (P 0.018)) and IM patients 25.276.1% (P 0.042), whereas o o EMs IMs PMs the 5-year estimate for relapse or OS was not statistically different between the groups. Four of the nine PM-patients Figure 2 Maximum bilirubin levels in mg/dl before and after transplant died from TRM between days 50 and 174 after transplant. for each study group. Poor metabolizers had the highest bilirubin and creatinine values after transplant compared to extensive metabolizers or Causes of death were multi-organ failure, including liver and intermediate metabolizers. kidney involvement in combination with an opportunistic infection (Aspergillus pneumonia n ¼ 1, bacterial pneumonia n ¼ 2and adenovirus infection n ¼ 1). Engraftment However, no differences in the 4-year estimates for Surprisingly, PM-patients had the fastest engraftment for TRM, relapse rate or OS were seen in patients with either neutrophils 4500 cells/ml with a median only 13 days EM, IM or PM donors (data not shown). Figure 3 shows (range 9–46) compared to IM-patients (median 16 days, TRM. range 9–33) or EM-patients (median 15 days, range 8–35),

Bone Marrow Transplantation Loss-of-function allele of CYP2C19 and transplant-associated mortality AH Elmaagacli et al 663 Estimates for treatment-related mortality versus others) on TRM defined as death other than relapse, was confirmed on a multivariate analysis considering all 1.0 P<0.017 PMs versus EMs potential factors which influence TRM: patient age, HLA- 0.9 P<0.045 PMs versus IMs constellation, sex-constellation between donor and patient, 0.8 T-cell depletion, disease stage, occurrence of acute GVHD. 0.7 Although many drugs used in the transplant setting may 0.6 PMs 50.0% + 18.6% induce exaggerated side effects in PMs and influence their 0.5 outcome, cyclophosphamide plays a distinctive role in this 0.4 setting with regard to toxicity. This cytostatic agent, widely 0.3 IMs 23.0% + 5.6% used in conditioning regimens in combination with TBI or BU, is metabolized by CYP enzymes.3,4 It is often 0.2 EMs 22.7% + 3.4% speculated that the broad interindividual variability of 0.1 cyclophosphamide’s side effects and clinical efficacy might 0.0 be due to differences in pharmacokinetics. Recently, 0123456789101112 cyclophosphamide-kinetics were analyzed for functionally Years post transplant relevant polymorphisms of the CYP metabolizing enzymes Figure 3 Estimates of transplant-related mortality. Poor metabolizers CYP2B6, CYP2C9, CYP2C19, CYP3A5 and glutathione- had the highest rate of TRM with 50% compared to extensive metabolizers S-transferase A1.3 The results revealed that the and intermediate metabolizers. CYP2C19*2 genotype (here defined as PMs) influenced pharmacokinetics of cyclophosphamide significantly, and but these differences were not statistically significant. thereby the serum concentrations of cyclophosphamide. Donor genotype also had no influence on time of Known side effects of cyclophosphamide are hepato- and engraftment (data not shown). nephrotoxicities. It is therefore possible that the slower metabolism of this cytostatic agent in PMs has an impact Multivariate analysis for TRM on the exaggerated TRM reported here. Six of the nine For TRM, age (p40 or 440 years), gender constellation PM-patients studied here received high doses of cyclopho- (male patient with female donor or other), loss-of-function sphamide as part of their conditioning regimen. In fact, of allele of CYP2C19 gene on the patient’s side (PM) versus the six patients receiving high doses of cyclophosphamide, others, HLA-constellation between patient and donor, three died from multi-organ failure. It might be helpful sibling donor versus unrelated donor, graft-type (PBSC to determine cyclophosphamide serum concentrations in or bone marrow), T-cell depleted graft versus non T-cell PM-patients in future to avoid increased side effects. depleted graft, and disease stage (advanced versus early Other prominent substrates of CYP2C19, which might disease stages), acute GVHD grade 2–4 were analyzed. potentially induce organ-toxicities and are widely used in Only the following factors: disease stage acute GVHD the transplant setting, include proton pump inhibitors (PPI) grade 2–4 (RR 2.7, 1.59–4.74; (95% CI)) (Po0.001), such as , or pantozole. Although disease stage (RR 2.6 1.41–4.86 (95% CI)) (Po0.01) and extensive toxicities of PPI on liver and kidney are not so loss-of-function allele of CYP2C19 gene (RR 2.0, 1.16– common, their side effects and thereby their influence on 3.31; (95% CI)) (Po0.012) influenced TRM. transplant outcome is not fully understood. It is reported in a different setting that their use in patients with different genotypes of CYP2C19 not only leads to varieties in Discussion measured gastric pH in these patients, but also causes differences in the eradication rate of helicobacter pylori This study provides evidence that allogeneic transplant- when the combination of lansoprazole, amoxicillin and ation in patients with a loss-of-function allele of CYP2C19 clarithromycin was used. Interestingly, the poorest eradica- gene is associated with increased transplant-related hepato- tion rate was found not in PMs, but in EMs.5–7 All of our and nephrotoxicities. Thus, we found a mean of twofold study-patients received omeprazole. higher maximum total bilirubin and maximum creatinine Moreover, little is known about organ toxicities caused levels in patients genotyped as PMs of CYP2C19 after by drug interactions involving metabolism of CYP2C19 transplant, compared to EMs with two active alleles, or enzymes. Some substrates of CYP2C19 are metabolized IMs with only one active allele for CYP2C19.In with a different affinity and by an alternative metabolic consequence, and as hypothesized, the increased organ pathway than other CYP enzymes. , a potent toxicity resulted in a markedly exaggerated TRM. Thus, drug, is such a , which is metabolized our calculated 4-year TRM-estimates for PMs increased up by CYP2C19 and has a 50 times lower affinity for CYP3A4 to 50%, which is also approximately twofold higher than enzymes. Coadministration of drugs with strong inhibitory those of EMs and IMs with about only 23% (Po0.02for effects on CYP3A4, such as ritonavir, a human immuno- PMs versus EMs, Po0.05 for PMs versus IMs). However, deficiency virus protease inhibitor, prevents metabolism of the current reported rate of TRM for EMs and IMs is in voriconazole by CYP3A4 enzymes and thereby leads to line with other studies concerning allogeneic transplants.19 higher voriconazole concentrations with the risk of Furthermore, we found that in PMs the major cause for increased adverse events such as hepatotoxicity.9 Indeed, death was multi-organ failure in combination with oppor- hepatotoxicity is a dose-limiting adverse event in patients tunistic infection. The impact of genotype CYP2C19 (PMs taking voriconazole, which is often observed. Moreover,

Bone Marrow Transplantation Loss-of-function allele of CYP2C19 and transplant-associated mortality AH Elmaagacli et al 664 clinically significant hepatic dysfunction occurs in more 1-week regimen of proton pump inhibitor, amoxicillin and than 10% of patients receiving regular doses of voricona- clarithromycin. Aliment Pharmacol Ther 2003; 17: 259–265. zole and is a major cause of discontinuation of this drug.20 7 Miki I, Aoyama N, Sakai T, Shirasaka D, Wambura CM, Studies conducted in Caucasian and Japanese healthy Maekawa S et al. Impact of clarithromycin resistence and subjects have shown that PMs of CYP2C19 have, on average, CYP2C19 genetic polymorphism on treatment efficacy of a fourfold higher exposure to voriconazole than do their Helicobacter pylori infection with lansoprazole- or rabepra- zole-based triple therapy in Japan. Eur J Gastroenterol Hepatol homozygous EM counterparts. Subjects who are heterozy- 2003; 15: 27–33. gous EMs experience, on average, a twofold higher vorico- 8 Kropp S, Lichtinghagen R, Winterstein K, Schlimme J, 21 nazole exposure than their homozygous EM counterparts. Schneider U. Cytochrome p-450 2D6 and 2C19 polymorph- Since many potentially CYP inhibiting drugs are often isms and length of hospitalization in psychiatry. Clin Lab 2006; used in the transplant setting such as metoclopramide, 52: 237–240. norfloxacin, clarithromycin, fluconazole or itraconazole, it 9 Mikus G, Scho¨ wel V, Drzewinska M, Rengelshausen J, Ding is possible that more adverse side events occur from drug R, Riedel KD et al. Potent cytochrome P4502C19 genotype- interactions than is currently recognized. related interaction between voriconazol and the cytochrome As in other studies, we too did not determine the P4503A4 inhibitor ritonavir. Clin Pharmacol Therap 2006; 80: CYP2C19 activity phenotype here, since previous studies 126–135. 10 Hulot J-S, Bura A, Villard E, Azizi M, Remones V, have already demonstrated that CYP2C19 genotyping Goyenvalle C et al. Cytochrome P4502C19 loss-of-function identifies more than 90% of PMs, indicating good polymorphism is a major determinant of respon- 22,23 genotype-phenotype agreement. siveness in healthy subjects. Blood 2006; 108: 2244–2247. However, no differences were assessed in the rate of 11 Ottinger HD, Ferencik S, Beelen DW, Lindemann M, Peceny hepato- and nephrotoxicities between EMs and IMs. R, Elmaagacli AH et al. Hematopoietic stem cell transplant- Furthermore, gene polymorphism of their donors including ation: contrasting the outcome of transplantations from donors typed as PMs had no influence on patient outcome. HLA-identical siblings, partially HLA-mismatched unrelated In conclusion, our study provides strong evidence that donors. Blood 2003; 102: 1131. TRM and morbidity are influenced by gene polymorphisms 12Elmaagacli AH, Peceny R, Steckel N, Trenschel R, Grosse- et al. of CYP2C19. Our results suggest that patients planned for Wilde H, Schaefer UW Outcome of transplantation of highly purified peripheral blood CD34+ cells with T-cell add- allogeneic transplantation should be genotyped for back compared with unmanipulated bone marrow or peri- CYP2C19 before transplant to identify those with a high pheral blood stem cells from HLA-identical sibling donors in individual risk of TRM. However, more studies concerning patients with first chronic phase chronic myeloid leukemia. the role of gene polymorphisms of CYP are necessary. Blood 2003; 101: 446. 13 Thomas ED, Storb R, Clift RA, Fefer A, Johnson L, Neiman PE et al. Bone marrow transplantation. N Engl J Med 1975; Acknowledgements 292: 895. 14 Maniatis T, Fritsch EF, Sambrook J. Molecular Cloning. A We thank Christiane Schary, Melanie Kroll, Silke Gottwald and Laboratory Manual. Cold Spring Harbor Laboratory Press: Ines Riepenhoff (all Essen, Germany) for their excellent Harbor, NY, USA, 1982. technical assistance with PCR analyses. This work was 15 Kaplan EL, Meier P. Non-parametric estimation from supported by grants from Deutsche Krebshilfe 70-3093-El4 incomplete observations. J Am Stat Assoc 1958; 53: 457–481. and Kulturstiftung Essen (05 032Elmaagacli). 16 Cox DR. Regression models and life-tables. J R Stat Soc 1972; 34: 187. 17 Luo H-R, Poland RE, Lin K-M, Wan YJY. Genetic polymorph- References ism of cytochrom P4502C19 in Mexican Americans: a cross- ethnic comparative study. Clin Pharmacol Ther 2006; 80: 33–40. 1 Daly AK. Pharmacogenetics of the major polymorphic 18 Halling J, Petersen MS, Damkier P, Nielsen F, Grandjean P, metabolizing enzymes. Fundam Clin Pharmacol 2003; 17: 27–41. Weihe P et al. Polymorphism of CYP2D6, CYP2C19, CYP2C9 2Desta Z, Zhao X, Shin JG, Flockhart DA. Clinical significance and CYP2C8 in the Faroese population. Pharmacogenetics 2005. of the cytochrome P4502C19 genetic polymorphism. Clin 19 Elmaagacli AH, Koldehoff M, Hindahl H, Steckel NK, Pharmacol Ther 2005; 77: 40. TrenschelR,PecenyRet al. Mutations in innate immune system 3 Timm R, Kaiser R, Lotsch J, Heider U, Sezer O, Weisz K et al. NOD2/CARD 15 and TLR-4 (Thr399Ile) influence the risk Association of cyclophosphamide pharmacokinetics to poly- for severe acute GVHD in patients who underwent an allogeneic morphic cytochrome P4502C19. J 2005; 5: transplantation. Transplantation 2006; 81: 247–254. 365–373. 20 Boucher HW, Groll AH, Chiou CC, Walsh TJ. Newer 4 Griskevicius L, Yasar U, Sandberg M, Hidestrand M, Eliasson systemic antifungal agents: pharmacokinetics, safety and E, Tybring G et al. Bioactivation of cyclophosphamide: the efficacy. Drugs 2004; 64: 1997–2000. role of polymorphic CYP2C enzymes. Eur J Clin Pharmacol 21 Ikeda Y, Umemura K, Kondo K, Sekiguchi K, Miyoshi S, 2003; 59: 103–109. Nakashima M. Pharmacokinetics of voriconazole and cyto- 5 Ishida Y, Goto Y, Kondo T, Kurata M, Nishio K, Kawai S chrome P4502C19 genetic status. Clin Pharmacol Ther 2004; et al. Eradication rate of Helicobacter pylori according to 75: 587–588. genotypes of CYP2C19, IL-1B, and TNF-A. Int J Med Sci 22 Wilkinson GR. Drug metabolism and variability among 2006; 3: 135–140. patients in drug response. N Engl J Med 2005; 352: 2211–2221. 6 Kawabata H, Habu Y, Tomioka H, Kutsumi H, Kabayashi N, 23 Andersson T, Flockhart DA, Goldstein DB, Goldstein DB, Oyasu K et al. Effect of different proton pump inhibitors, Huang SM, Kroetz DL et al. Drug-metabolizing enzymes: differences in CYP2C19 genotype and resistance on evidence for clinical utility of pharmacogenomic tests. Clin the eradication rate of Helicobacter pylori infection by a Pharmacol Ther 2005; 78: 559–581.

Bone Marrow Transplantation