An Evaluation of Engraftment, Toxicity and Busulfan Concentration in Children Receiving Bone Marrow Transplantation for Leukemia Or Genetic Disease

Home , Busulfan

Bone Marrow Transplantation (2000) 25, 925–930  2000 Macmillan Publishers Ltd All rights reserved 0268–3369/00 $15.00 www.nature.com/bmt An evaluation of engraftment, toxicity and busulfan concentration in children receiving bone marrow transplantation for leukemia or genetic disease

AM Bolinger1, AB Zangwill1, JT Slattery2,3, D Glidden4, K DeSantes5, L Heyn6, LJ Risler3, B Bostrom7 and MJ Cowan8

1University of California at San Francisco, Department of Clinical Pharmacy; 2Department of Pharmaceutics, University of Washington; 3Fred Hutchinson Cancer Research Center; 4University of California at San Francisco, Department of Epidemiology and Biostatistics; 5University of Wisconsin, Department of Pediatrics; 6University of California at San Francisco, Department of Nursing; 7University of Minnesota, Department of Pediatrics; and 8University of California at San Francisco, Department of Pediatrics, San Francisco, CA, USA

Summary: children exhibit average busulfan steady-state concentrations (Css) or an area under the curve (AUC) substan- Autologous recovery is a major problem with busulfan tially less than do older children or adults.7–9 The Css corre- as a marrow ablative agent in conditioning children for sponds to the AUC over a dosing interval divided by the allogeneic BMT. Data suggest the average concentration 6 h between doses. The AUC over a dosing interval is equal of busulfan at steady state (Bu Css) is critical for suc- to the AUC from the time the first dose is ingested to infin- cessful engraftment. We prospectively evaluated busul- ity if no subsequent doses are taken. A Css of 900 ng/ml fan pharmacokinetics in 31 children (age 0.6–18 years) is equal to an AUC of 1350 ␮m × min. with AML (n = 9), and non-malignant diseases (n = 22) Few studies of the relationship between busulfan receiving HLA-closely matched (sibling, parent, Css (Bu Css) or AUC and outcome of the busulfan/ unrelated) donor grafts. Blood samples were obtained cyclophosphamide preparative regimen have focused on following dose 1 and 13 of a standard 16 dose, 4-day children. Adults with CML have been shown to be at regimen. The busulfan dose varied from 14 to 20 mg/kg. increased risk of relapse if Bu Css is low and exhibit less Patients received cyclophosphamide 200–240 mg/kg; busulfan-related toxicity compared to patients with other 22/31 received 80–90 mg/kg of ATG. Eight patients diseases receiving similar treatment.10,11 Excessively low failed to engraft (26%). ATG did not appear to influ- = Ͻ levels of busulfan have been related to graft rejection, while ence engraftment (P 0.38). Bu Css levels 600 ng/ml high levels have been related to severe toxicities.12,13 Of correlated with autologous recovery/mixed chimerism particular interest in children is the issue of graft rejection. (P = 0.018). There were no graft failures in patients with Since cyclophosphamide is directly toxic to mature lym- a Bu Css Ͼ600 ng/ml. A correlation between Bu Css phocytes and busulfan is not, graft rejection has been levels and regimen-related toxicity (RRT) was not ident- thought to be due to inadequate dosing of cyclophospham- ified for grade 2 or higher toxicities, only 1/31 had a Bu 1,14 Css Ͼ900 ng/ml. Our data support the use of pharmaco- ide. Thus, the relatively high incidence of graft rejection kinetic monitoring of busulfan. Bone Marrow Transplan- in children receiving a busulfan/cyclophosphamide prepara- tation (2000) 25, 925–930. tive regimen has resulted in the use of higher doses of Keywords: busulfan; children; bone marrow transplan- cyclophosphamide than are used in adults. ATG also has tation; cyclophosphamide; toxicity been added to the regimen in an attempt to gain additional immune suppression. However, graft rejection is still relatively frequent, even with total doses of cyclophosphamide as high as 200–240 mg/kg and ATG. We therefore sought Busulfan is a potent myelotoxin that complements the lym- to determine whether graft rejection was associated with phocyte toxicity of cyclophosphamide in preparative regi- low Bu Css in children receiving 14–20 mg/kg busulfan mens for hematopoietic stem cell transplantation.1–3 The and 200–240 mg/kg cyclophosphamide, whether toxicities standard dose is 1 mg/kg administered orally every 6 h over (primarily to the gastrointestinal (GI) mucosa, liver and 4 days. However, this regimen results in highly variable lung) were associated with high busulfan levels and concentrations in plasma, particularly when young children whether ATG facilitated engraftment in our patients who are compared to older children or adults.4–7 At a given dose, were conditioned with busulfan and cyclophosphamide. normalized to either body weight or surface area, young

Correspondence: Dr AM Bolinger, University of California, San Franci- sco, School of Pharmacy, 521 Parnassus, Room C-152 Box 0622, San Francisco, CA 94143-0622, USA Received 18 October 1999; accepted 19 January 2000 Busulfan concentration and BMT outcome in children AM Bolinger et al 926 Methods ning at 9 a.m. on the ﬁrst day of conditioning. In the 31 patients evaluated, the total oral busulfan dose varied from Patients and treatments 14 to 20 mg/kg. Busulfan was followed by cyclophosphamide administered i.v. at 50 to 60 mg/kg on 4 consecutive We prospectively evaluated busulfan pharmacokinetics in days (only one patient received a total dose of 240 mg/kg); n = 38 pediatric patients (age 0.6 to 18 years) with AML ( 22/31 patients (71%) received a total dose of 80 to 90 14) and non-malignant diseases (n = 24) who were receiv- mg/kg ATG (Table 1). A compounded suspension of oral ing HLA-matched or closely matched sibling (n = 28), par- busulfan was administered through a nasogastric tube in n = n = ent ( 3), or unrelated ( 4) donor marrow grafts. Donor children unable to reliably swallow tablets. All other cell dose was approximately 5 × 108 per kilogram for all patients received doses as commercially available 2 mg tab- patients. Patient demographics are described in Table 1. All lets. Patients were NPO for at least 1 h before and 1 h after patients enrolled in BMT protocols from May 1991 to April busulfan administration. Graft-versus-host disease (GVHD) 1996 that included busulfan as a conditioning agent were prophylaxis was methotrexate (MTX) (n = 1); cyclosporine eligible. Seven patients were excluded: two received auto- (CsA) (n = 3); or a combination of CsA/MTX (n = 25), logous transplants, one received a syngeneic transplant, two CsA/MTX/methylprednisolone (n = 3) or CsA/prednisone received a T cell-depleted transplant, one received mel- (n = 1). Phenytoin (PHT), dose adjusted to maintain thera- phalan and one was excluded due to problems in obtaining peutic levels, was administered to nine patients (29%) as samples. Parents of all patients provided informed consent prophylaxis against busulfan-related seizures. for participation in this study which was approved by the Committee on Human Research at the University of Cali- fornia, San Francisco. Patients over 12 years of age also Samples and pharmacokinetic analysis provided assent to participate. Busulfan was administered as part of the BMT prepara- Blood samples were collected for analysis in heparinized tive regimen as the standard 16 dose, 4 day regimen begin- tubes at times 0, 30, 60, 120, 180, 240, 300 and 360 min

Table 1 Patient demographics

UPN Age (years) Diagnosis Donora Weight (kg) Bu/Cy dose ATG (mg/kg) GVHD Other (mg/kg) prophylaxis

123 0.9 AML 1 8.7 16/200 0 MTX 124 1.9 Hurlers 2 13.2 16/200 80 MTX/CsA 126 0.8 SCID 1 8 16/200 80 MTX/CsA 131 2 AML 1 10.9 16/200 0 CsA 143 8.2 ALD 1 22.7 18/240 80 MTX/CsA 146 6.8 B-Thalassemia 1 19.5 16/200 80 MTX/CsA 152 1.3 AML 1 12.9 16/200 0 CsA 153 0.8 B-Thalassemia 1 9.6 16/200 80 MTX/CsA 155 2.9 B-Thalassemia 1 14.6 16/200 80 MTX/CsA 156 12 B-Thalassemia 1 24 16/200 80 MTX/CsA 165 12.5 EPHL 1 45.7 16/200 0 MTX/CsA VP-16/PHT 170 12 BFD 1 32 16/200 80 MTX/CsA PHT 175 17 AML 1 82.7 16/200 0 MTX/CsA 181 2.7 B-Thalassemia 1 16 16/200 80 MTX/CsA 188 10 B-Thalassemia 1.4 30.3 16/200 80 MTX/CsA 191 2.2 WAS 1.3 12.9 16/200 80 MTX/CsA 195 12.3 Sickle cell anemia 1 54b 14/200 90 MTX/CsA PHT 198 17.9 AML 1 42.6 16/200 0 MTX/CsA PHT 199 0.7 WAS 2 7.6 20/200 80 MTX/CsA/MP PHT 205 5.9 BFD 1 15 16/200 80 MTX/CsA 208 5.3 B-Thalassemia 1 17 16/200 80 MTX/CsA 209 7.8 B-Thalassemia 1 20 16/200 80 MTX/CsA 214 11.7 AML 1 42 16/200 0 MTX/CsA 215 3.5 B-Thalassemia 1 17.6 16/200 80 MTX/CsA 218 1.6 Hurlers 2 12.1 20/200 80 MTX/CsA/MP PHT 223 2.9 AML 2 14 16/200 80 MTX/CsA/MP 229 1.1 B-Thalassemia 1 11.2 16/200 80 MTX/CsA PHT 232 0.7 CID 1 8 16/200 80 MTX/CsA 239 4.9 AML 1 17.9 16/200 0 MTX/CsA 241 16.3 AML 1 62 16/200 0 CsA/PDN PHT 252 6.2 Sickle cell anemia 1 18.6 16/200 90 MTX/CsA PHT

UPN = unique patient number; Bu = busulfan; CY = cyclophosphamide; ATG = anti-thymocyte globulin; GVHD = graft-versus-host disease; AML = acute myelogenous leukemia; SCID = severe combined immunodeﬁciency syndrome; ALD = adrenoleukodystrophy; EPLH = erythrophagocytic lymphohi- stiocytosis; BFD = Blackfan–Diamond syndrome; WAS = Wiskott–Aldrich syndrome; CID = combined immunodeﬁciency disease; MTX = methotrexate; CsA = cyclosporine; MP = methylprednisolone; PDN = prednisone; PHT = phenytoin. aDonor key: 1 = HLA matched sib; 1.1 = mismatched sib; 1.3 = mismatched parent; 1.4 = matched parent; 2 = HLA matched URD. bChild’s weight adjusted for obesity.

Bone Marrow Transplantation Busulfan concentration and BMT outcome in children AM Bolinger et al 927 following the first and thirteenth busulfan doses. Samples (89%) patients who did not receive ATG. The difference were stored on ice until the plasma was separated and (P = 0.38) was not statistically significant. frozen at −20°C until analysis. Specimens were sent on dry The use of mean Bu Css as a predictor condenses two ice by overnight express mail for analysis by gas measurements into one. This may result in the loss of pre- chromatography/mass spectrometry. Samples from the first dictive information. For instance, a subject with the levels 12 patients were assayed at the University of Minnesota. of 400 ng/ml and 800 mg/ml has the same average level All remaining samples were assayed by the Fred Hutchin- as a subject with 550 ng/ml and 650 ng/ml, respectively. son Cancer Research Center (FHCRC).15 Bu Css was calcu- Such differences may have implications for the probability lated as previously described.13 Data reported for Bu Css of engraftment. To examine this, we fitted the model to are the mean value for each patient. the 24 subjects with two blood level measurements. Two predictors were included in a logistic regression model; the average blood level and the difference between the average Toxicity and rejection and the minimum blood level. The model found strong pre- Regimen-related toxicity (RRT) was evaluated using BMT dictive value in the former but no predictive information criteria described by Bearman et al16 during the first 28 in the latter variable. Hence, it appears that the probability days post transplant and assigned a grade of 1 to 4. of engraftment depends on Bu Css rather than the minimum Engraftment was defined as an absolute ANC Ͼ500 for 3 or maximum exposure. consecutive days with evidence of donor cells either by The incidence of graft rejection in relation to busulfan PCR for non-HLA DNA polymorphisms and/or in situ concentration and source of graft is described in Table 3. hybridization for X and Y chromosomes in sex-mismatched Of the 31 evaluable patients, eight failed to engraft (two were mixed chimeras), giving a 74% successful full donor transplants. A quantitative estimate of engraftment (percent Ͻ donor cells) was provided for sex-mismatched transplants. engraftment rate. Bu Css levels 600 ng/ml significantly correlated with autologous recovery/mixed chimerism For non-HLA DNA assays, unique donor and recipient = alleles were identified to assess full donor chimerism according to the Wilcoxon rank-sum test (P 0.018). There were no graft failures in the eight patients with Bu Css (sensitivity of 1–5%). For analysis, we included as Ͼ ‘engrafted’ only those patients with 100% donor chimer- 600 ng/ml. The logistic regression model was used to fit ism. Primary graft rejection was defined as no evidence of a dose–response relationship between average Bu Css and donor cells; secondary rejection was documented by initial probability of engraftment in 30 subjects (Figure 1). An evidence of donor engraftment followed by full autologous alternative approach was to use a model-free scatterplot recovery. Engraftment was evaluated at 3 month intervals smoother (Lowess) to develop a dose response relationship up to 1 year post BMT. (Figure 1). The logistic model predicts that the Bu Css that ensures a 90% chance of engraftment is 639 ng/ml and that 762 ng/ml is required to ensure a 95% chance of Statistical analysis engraftment. The Lowess smoother predicts that 90% of subjects will successfully engraft with an average Bu Css Average Bu Css between engrafted and non-engrafted of 532 ng/ml and 95% of subjects will be engrafted success- patients was compared by Wilcoxon rank-sum test. The fully with a Bu Css of 569 ng/ml. While the data show that logistic regression model17 was used to relate Bu Css, donor all eight subjects who were evaluated with levels greater source, GVHD prophylaxis and other patient characteristics than 600 ng/ml successfully engrafted, we do not have (weight, age, cyclophosphamide and ATG) to probability enough data to clearly identify the threshold Bu Css at of engraftment or toxicity. Donor source and GVHD which engraftment is ensured. Of the eight patients who prophylaxis were analyzed with relation to engraftment failed to engraft, seven had genetic disorders and one had using the Fisher exact test. A scatter plot smoother18 was AML. While there does not appear to be a correlation used to develop a model-free estimate on the average between diagnosis and engraftment, our numbers are too relationship between engraftment probability and Bu Css. small to reach a conclusion.

Results Busulfan concentration and toxicity

Busulfan concentration and graft rejection All 38 patients were included in the toxicity analysis (Table 4). A high percentage of patients (36/38) experienced grade Busulfan concentration and transplant outcome are shown 1 GI toxicity (diarrhea) which may have been in part caused in Table 2. The mean busulfan level was deﬁned as the by GI antimicrobial decontamination with oral vancomycin average of the ﬁrst and second values when available. For and gentamicin. Grade 1 or 2 hepatic toxicity was observed subjects with only one value, the mean listed is that value. in 30 patients (79%) and 14 (37%) experienced grade 2 The population average of the mean blood levels was 528 or less renal toxicity. There were four patients (11%) with ng/ml. The median was 473 ng/ml with a range of 157– seizures (grade 3 neurotoxicity) attributed to the condition- 2061 ng/ml. The subject with a single outlying value of ing regimen; none of these children received seizure 3543 ng/ml was included in statistical analysis but excluded prophylaxis. Since seizures are felt to be a preventable tox- from the curves shown in Figures 1 and 2. Of patients who icity, these patients were not included in the evaluation of received ATG, 15/23 (65%) engrafted compared with 8/9 Bu Css and risk of toxicity. There were two cases (5%) of

Bone Marrow Transplantation Busulfan concentration and BMT outcome in children AM Bolinger et al 928 Table 2 Busulfan concentration and transplant outcome

UPN Bu Css (ng/ml) Overall RRT Major RRT Statusa

Dose 1 Dose 13 Mean

123 925 335 630 1 hepatic 1 124 334 N/A 334 2 hepatic 0 126 428 588 508 2 stomatitis 1 131 157 N/A 157 2 renal 0 143 579 3543 2061 3 CNS 1 146 703 243 473 1 hepatic/GI 1 152 154 200 177 1 hepatic/GI 1 153 792 345 565 1 hepatic/GI 1 155 322 62 192 2 hepatic 1 156 378 89 233 2 hepatic 0 165 N/A 468 468 2 stomatitis 1 170 N/A 423 423 4 hepatic 1 175 532 N/A 532 2 hepatic/stomatitis 1 181 N/A 463 463 1 hepatic/GI 1 188 579 626 602 1 hepatic/GI 1 191 506 497 502 2 hepatic 1 195 526 710 618 1 renal/GI 1 198 813 617 715 3 bladder 1 199 327 N/A 327 1 GI 1 205 333 349 341 1 hepatic/GI 1 208 379 466 422 1 hepatic/GI 0 209 624 503 563 2 stomatitis 1 214 673 648 660 3 CNS 1 215 549 594 571 2 renal/hepatic 0 218 574 321 448 2 hepatic 0 223 595 894 745 3 CNS 1 229 497 360 429 2 hepatic 0 232 352 426 389 1 hepatic/GI 0 239 581 557 569 2 renal 1 241 1011 658 835 1 renal/hepatic/GI 1 252 520 332 426 1 renal/GI 1

UPN = unique patient number; Bu = busulfan; RRT = reimen-related toxicity; CNS = central nervous system; GI = gastrointestinal system; N/A = not available. aStatus key: 1 = engrafted; 0 = not engrafted.

100 100

80 80

60 60

40 Logistic 40 Lowness 20 20 Probability of Engraftment 0 0 200 400 600 800 Fraction with grade 3 or higher toxicity 200 400 600 800 Average BU Css ng/ml Average BU ss ng/ml Figure 1 Probability of engraftment compared to Bu Css (ng/ml). Figure 2 Fraction of patients with toxicity as related to Bu Css (ng/ml).

hemorrhagic cystitis (grade 3 bladder toxicity) and one case of fatal VOD (grade 4 hepatic toxicity). higher toxicity. The data show that only a single patient The logistic regression model was again employed to with levels Ͻ600 ng/ml had a grade 3 or higher toxicity. relate the risk of toxicity to Bu Css in 30 patients. The Among the seven patients evaluated with blood levels data show virtually no dose–response relationship between у600 ng/ml, three (42%) developed a grade 3 or higher blood level and risk of grade 2 or higher toxicity. Figure toxicity, two of which were seizures. The logistic model 2 shows the ﬁt of the model to the risk of a grade 3 or predicts that this probability may vary considerably over

Bone Marrow Transplantation Busulfan concentration and BMT outcome in children AM Bolinger et al 929 Table 3 Incidence of graft rejection in relation to busulfan concen- toxicity with Bu Css Ͼ900 ng/ml.7 Our results differ from tration and source of graft these previous studies in that we observed severe RRT (Ͼgrade 2) in 2/31 evaluated patients with no dependence Source of graft Busulfan average steady-state P value on Bu Css. The principal difference between our study and concentration (ng/ml) Ͻ600 vs Ͼ600 those just cited is that the dose of cyclophosphamide was higher in our pediatric patients than in adult patients in Ͻ200 200–600 Ͼ600 whom a relationship between Bu Css and RRT has been Overall 1/3 7/20 0/8 0.018 previously reported. Since only one of our patients had a HLA partially-matched Bu Css Ͼ900 ng/ml (this patient had a mean level of 2061 related or URD 0/0 2/3 0/1 ng/ml and a grade 3 CNS RRT), we are unable to comment Matched sibling 1/3 5/17 0/7 on the incidence of toxicity above this previously defined threshold. Our data suggest that higher Bu Css levels (Ͼ600 ng/ml) confer increased risk for RRT, but due to the small number of patients, statistical significance could not be established. Table 4 Regimen-related toxicity Vassal et al8 conducted a study in 27 children with malignant tumors using busulfan 600 mg/m2, a dose felt to Grade 1 Grade 2 Grade 3 Grade 4 achieve AUCs comparable to adults receiving standard dose therapy (16 mg/kg). This was substantiated by the fact that Heart 0 0 0 0 Bladder 6 0 2 0 extra-hematologic toxicities approached those seen in the Mucosa 3 9 0 0 adult population. The authors specifically found an increase GI tract 36 0 0 0 in neurotoxicity from 1.7% to 15.4%. None of these Kidney 8 6 0 0 patients received phenytoin prophylaxis. This, along with Hepatic 20 10 0 1 Lung 0 0 1 0 our observation of grade 3 neurotoxicity in four patients CNS0040not receiving phenytoin, supports the need for seizure prophylaxis in all children receiving busulfan-containing conditioning regimens. Tutschka et al21 reported a decrease in the incidence of RRT when the total dose of cyclophosphamide was lowered this range. A threshold for grade 3 and 4 toxicity was to 120 mg/kg from 200 mg/kg in adults. In contrast, the not identified. study cited above by Pawlowska et al20 where 27/64 patients received 120 mg/kg cyclophosphamide, severe RRT was observed at relatively low busulfan levels and Discussion there was no relationship between RRT and busulfan AUC. In summary, we found that subjects with Bu Css Ͼ600 The major findings of this investigation are that ng/ml have a high probability of engraftment, that grade 2 engraftment following hematopoietic stem cell transplan- or higher toxicity does not appear to be related to Bu Css, tation using a busulfan and cyclophosphamide preparative and grade 3 or 4 toxicities may be related to Bu Css. Use regimen correlates with Bu Css while the incidence of regi- of lower dose cyclophosphamide (120 mg/kg) may help men-related toxicity did not. decrease RRT while maintaining adequate immunosuppres- Relationships between Bu Css and graft rejection have sion. Our data support the use of targeted pharmacokinetic been reported previously. Based on earlier work, thresholds monitoring to maintain Bu Css above 600 ng/ml to maxim- for rejection appear to be 200 ng/ml for HLA matched and ize the rate of successful engraftment. The data also suggest partially matched grafts from related donors.13,19 Our that a trial of targeted Bu Css with lower cyclophosphamide results differ in that we find no dependence of rejection on doses is warranted. the source of graft. In addition, a study by Pawlowska et al20 in pediatric ␤-thalassemia patients receiving HLA- matched sibling grafts after 14–16 mg/kg busulfan with Acknowledgements 120–200 mg/kg cyclophosphamide did not find an association between graft rejection and busulfan AUC (rejection This work was supported in part by NIH grant M01-RR01271, was observed in 5/64 patients). We observed no rejections Pediatric Clinical Research Center; NIH grant CA18029, Chil- with Bu Css Ͼ600 ng/ml regardless of donor recipient dren’s Cancer Research Fund of the University of Minnesota. match. In busulfan- and cyclophosphamide-containing regimens, three studies have reported a threshold for the development References of severe RRT in adults with Bu Css Ͼ900–1000 ng/ml (which corresponds to an AUC of 1350–1500 ␮M ϫ 11–13 1 Tutschka PJ, Santos GW. Bone marrow transplantation in the min). An exception is the report on adult CML patients busulfan-treated rat: relationship between myelosuppression receiving HLA-matched sibling grafts following condition- and immunosuppression for conditioning bone marrow recipi- ing with 16 mg/kg busulfan and 120 mg/kg cyclophos- ents. Transplantation 1977; 24: 52–62. phamide. In these patients there was no increased risk of 2 Santos GW, Tutschka PJ. Marrow transplantation in the busul-

Bone Marrow Transplantation Busulfan concentration and BMT outcome in children AM Bolinger et al 930 fan treated rat: preclinical model of aplastic anemia. J Natl 12 Dix SP, Wingard JR, Mullins RE et al. Association of busul- Cancer Inst 1974; 53: 1781–1785. fan area under the curve with veno-occlusive disease follow- 3 Peters WP, Henner WB, Growchow LB et al. Clinical and ing BMT. Bone Marrow Transplant 1996; 17: 225–230. pharmacologic effects of high dose single agent busulfan with 13 Growchow LB. Busulfan disposition: the role of therapeutic autologous bone marrow support in the treatment of solid monitoring in bone marrow transplantation induction regi- tumors. Cancer Res 1987; 47: 6402–6406. mens. Semin Oncol 1993; 20: 18–25. 4 Grochow LB, Krivit W, Whitley CB, Blazer BR. Busulfan 14 Santos GW, Tutschka PJ, Brookmeyer R et al. Marrow trans- disposition in children. Blood 1990; 75: 1723–1727. plantation for acute nonlymphocytic leukemia after treatment 5 Hassan M, Oberg G, Bekassy AN et al. Pharmacokinetics of with busulfan and cyclophosphamide. New Engl J Med 1983; high dose busulphan in relation to age and chronopharma- 309: 1347–1353. cology. Cancer Chemother Pharmacol 1991; 28: 130–134. 15 Slattery JT, Risler LJ. Therapeutic monitoring of busulfan in 6 Vassal G, Gouyette A, Hartmann O et al. Pharmacokinetics of hematopoietic stem cell transplantation. Ther Drug Monit high dose busulfan in children. Cancer Chemother Pharmacol 1998; 20: 543–549. 1989; 24: 386–390. 16 Bearman SI, Applelbaum FR, Buckner CD et al. Regimen- 7 Gibbs JP, Murray G, Risler L et al. Age-dependent tetrahydro- related toxicity in patients undergoing bone marrow transplan- thiophenium ion formation in young children and adults tation. J Clin Oncol 1988; 6: 562–568. receiving high-dose busulfan. Cancer Res 1997; 57: 5509– 17 Agresti A. An Introduction to Categorical Data Analysis. 5516. Wiley: New York, 1996. 8 Vassal G, Deroussent A, Challine D et al. Is 600 mg/m2 the 18 Cleveland WS. Robust locally weighted regression and smoo- appropriate dosage of busulfan in children undergoing bone thing scatter plots. J Am Stat Assoc 1979; 74: 829–836. marrow transplantation? Blood 1992; 79: 2475–2479. 19 Regazzi MB, Locatelli F, Buggia I et al. Disposition of high- 9 Hassan M, Ljungman P, Bolme P et al. Busulfan bioavail- dose busulfan in pediatric patients undergoing bone marrow ability. Blood 1994; 84: 2144–2150. transplantation. Clin Pharmacol Ther 1993; 53: 45–52. 10 Slattery JT, Clift RA, Buckner CD et al. Marrow transplan- 20 Pawlowska AB, Blazar BR, Angelucci E et al. Relationship tation for chronic myeloid leukemia: the inﬂuence of plasma of plasma pharmacokinetics of high-dose oral busulfan to the busulfan levels on the outcome of transplantation. Blood 1997; outcome of allogeneic bone marrow transplantation in children 89: 3055–3060. with thalassemia. Bone Marrow Transplant 1997; 20: 915– 11 Slattery JT, Sanders JE, Buckner CD et al. Graft rejection and 920. toxicity following bone marrow transplantation in relation to 21 Tutschka PJ, Copelan EA, Klein JP. Bone marrow transplan- busulfan pharmacokinetics. Bone Marrow Transplant 1995; tation for leukemia following a new busulfan and cyclophos- 16: 31–42. phamide regimen. Blood 1987; 70: 1382–1388.

Bone Marrow Transplantation