Low plasma busulfan (Bu) area under the concentration-time curve (AUC) is associated with graft failure and relapsed leukemias, and high AUC with toxicities when Bu is used orally or i.v. 4 times daily combined with cyclophosphamide in myeloablative hematopoietic stem cell transplantation (SCT) conditioning regimens. We report Bu AUC and its association with clinical outcomes in 130 patients with hematologic malignancies given a once-daily i.v. Bu (3.2 mg/kg days -5 to -2) and fludarabine (Flu, 50 mg/m(2) days -6 to -2) regimen. Total-body irradiation (TBI) 200 cGy x 2 was added for 51 patients with acute leukemias. Plasma AUC varied 3.6-fold (2184-7794 microM.min, median 4699 microM.min). Patients with an AUC >6000 microM.min had lower overall survival (OS) than those with AUC < or =6000 microM.min at 12 months (38% versus 74%) and 36 months (23% versus 68%, P < .001). This effect was apparent in patients with standard-risk and high-risk disease, and persisted when potential confounders were considered (hazard ratio 3.2, 95% confidence interval 1.7-6.3). Nonrelapse mortality (NRM) at 100 days (6% versus 19%) and progression free survival (PFS; 58% versus 16%) at 3 years were better with AUC < or =6000 microM.min. These data support a role for therapeutic dose monitoring and dose adjustment with daily i.v. busulfan.

Busulfan is the only agent used in myeloablative regimens for hematopoietic stem cell transplantation for which therapeutic drug monitoring (TDM) has been widely used. Studies of oral busulfan (Bu) indicate wide intrapatient and interpatient variations in pharmacokinetic (PK) behavior, particularly in children. Dose adjustments of oral Bu based on TDM to bring exposures within established therapeutic ranges have been shown to reduce toxicity and improve outcomes. Intravenous (IV) Bu is becoming more widely used and has much more predictable PK. Outcomes with IV Bu appear to be superior to those achieved using oral Bu without TDM. However there is still at least a threefold variation in exposures achieved by the same dose of IV Bu in different individuals and a small proportion of patients will experience toxic exposures with current dosing regimens. Therapeutic monitoring with appropriate dose adjustment is therefore recommended for all patients treated with regimens containing high doses of Bu. Giving IV Bu at a fixed rate to adults will narrow the range of exposures but more work is needed to establish the best dosing regimen to bring as many exposures as possible within the target range. Studies of test dosing of IV Bu show that this strategy is more accurate when test and treatment doses are infused at the same rate. Finally, targeting exposures to the upper end of the therapeutic range may provide a safe approach to exploiting dose-intensity for the treatment of some malignancies.

Intravenous (i.v.) busulfan (Bu) administered once daily in myeloablative transplant regimens is convenient, effective, and relatively well tolerated. Therapeutic drug monitoring is recommended as nonrelapse mortality increases when daily exposure, as determined by the area under the plasma concentration versus time curve (AUC), exceeds 6000 μM·min. We describe sequential studies to achieve accurate prediction of treatment doses of Bu based on the kinetics of a smaller test dose. A total of 335 patients with hematologic malignancies were given daily i.v. Bu 3.2 mg/kg × 4 and fludarabine 50 mg/m(2) × 5. Pharmacokinetic monitoring was conducted for both the test dose and first treatment dose of Bu (day -5). Three different test dose schedules were evaluated: 12 mg Bu administered over 20 minutes, 0.8 mg/kg over 3 hours, and 0.8 mg/kg infused at 80 mg/h. The 3.2 mg/kg treatment doses were infused over a fixed time of 3 hours for the first 2 test dose trials and at a fixed rate of 80 mg/h for the final protocol. All test dose infusions were on day -7. In the first 2 schedules, Bu administered over a fixed time had significantly higher clearance for the test dose compared with the treatment dose. However, when both the test and the treatment doses were administered at the same infusion rate, clearance of the drug between the 2 dosing days was equivalent. Predicted day -5 AUC (AUC(-5)) showed a high linear correlation (r(2) = 0.74) to the actual AUC(-5). The error of these predictions was<20% in 98% of patients and <10% in 80%. In 24 individuals, the test dose predicted an AUC >5500 μM·min; therefore, the first Bu treatment dose was reduced to a desired target AUC. All adjusted doses fell within 20% of the targeted exposure. We conclude that a test dose strategy for therapeutic drug monitoring of daily i.v. Bu is accurate if the test and treatment doses are infused at the same rate. This approach allows targeting of therapeutic doses of Bu to desired levels and the potential for improved safety and efficacy.

A combination of fludarabine (Flu) and daily i.v. busulfan (Bu) is well tolerated and effective in patients undergoing allogeneic hematopoietic stem cell transplantation. Although there is some evidence that Bu exposures exceeding 6000 μM.min [corrected] may lead to excessive toxicity, there is little information on the effect of exposures below this level on outcomes. We studied Bu exposure, as measured by area under the concentration-time curve (AUC), in 158 patients with various hematologic malignancies in an attempt to identify an optimal range for targeted therapy. The preparative chemotherapy regimen comprised Flu 50 mg/m(2) on days -6 to -2 and i.v. Bu 3.2 mg/kg on days -5 to -2 inclusive. Graft-versus-host disease (GVHD) prophylaxis included methotrexate, cyclosporin A, and antithymocyte globulin. Patients with Bu exposures below the median AUC of 4439 μM.min [corrected] were at increased risk for acute GVHD grade II-IV (hazard ratio [HR], 2.30; 95% confidence interval [CI], 1.19 to 4.49; P = .014). Those in the highest and lowest Bu exposure quartiles (daily AUC <3814 μM.min and >4993 μM.min) [corrected] had an increased risk of nonrelapse mortality (subdistribution HR, 3.32; 95% CI, 1.46 to 7.54; P = .004), as well as worse disease-free survival (HR, 1.81; 95% CI, 1.09 to 2.99; P = .021) and overall survival (HR, 1.94; 95% CI, 1.12 to 3.37; P = .018). Bu exposures between 4440 and 4993 μM/min were accompanied by the lowest risk of both nonrelapse mortality and acute GVHD.

A B S T R A C T Antithymocyte globulin (ATG) levels and clearance vary significantly among patients receiving the same weightbased dose of ATG. To date, ATG area under the curve (AUC), its determinants, and its impact on clinical outcomes have been examined in pediatric hematopoietic cell transplant (HCT) and adult nonmyeloablative HCT. Here we set out to examine ATG AUC in 219 uniformly treated adults undergoing myeloablative allogeneic HCT at our institution. Sera were collected for the determination of pre-or post-HCT ATG AUC. The lowest quintiles of pre-and post-HCT AUC were associated with inferior chronic graft-versus-host disease (GVHD) and relapse-free survival (cGRFS) and a higher risk of acute GVHD, respectively. The highest pre-or post-HCT ATG AUC quintiles were not associated with risk of death, nonrelapse mortality, or relapse. Factors most strongly associated with AUC were day À2 recipient absolute lymphocyte count, body mass index (BMI), and graft lymphocyte content. To achieve ideal pre-HCT AUC (avoiding low AUC to maximize cGRFS) in this HCT setting, ATG dosing will need to take into consideration recipient weight, BMI, and blood and graft lymphocyte counts. Further studies are required to develop a modern ATG dosing schema and to demonstrate that adjusting ATG dose to target a particular AUC is feasible and leads to improved outcomes.

A physiologically based pharmacokinetic (PBPK) model of the DNA-alkylating agent busulfan was slightly modified and scaled from adults to children in order to predict the systemic busulfan drug exposure in children. Capitalizing on the recent major software release of PK-Sim®, we refined our PBPK model by implementing glutathione S transferase (GST) in 11 organs using the software integrated enzyme expression database. In addition, two irreversible binding processes (i.e., DNA and plasma protein binding) were applied by using Koff and KD values. The model was scaled from adults to children. Simulations were computed and compared to concentration-time data after intravenous (i.v.) busulfan administration to 36 children. Based on the results, an age-dependent enzyme activity and maturation ratio was tailored and evaluated with an external dataset consisting of 23 children. Initial adult to children scaling indicated lower clearance values for children in comparison to adults. Subsequent age-dependent maturation ratio resulted in three different age groups: Activity of busulfan-glutathione conjugate formation was 80%, 61%, and 89% in comparison to adults for children with an age of up to 2 years, > 2-6 years, and > 6-18 years, respectively. Patients of the evaluation dataset were simulated with a mean percentage error (MPE) for all patients of 3.9% with 3/23 children demonstrating a MPE of > ±30%. The PBPK model parameterization sufficiently described the observed concentration-time data of the validation dataset while showing an adequate predictive performance. This PBPK model could be helpful to determine the first dose of busulfan in children.

PurposeTamoxifen is a key therapeutic option for breast cancer treatment. Understanding its complex metabolism and pharmacokinetics is important for dose optimization. We examined the possibility of utilizing archival formalin-fixed paraffin-embedded (FFPE) tissue as an alternative sample source for quantification since well-annotated retrospective samples were always limited.MethodsSix 15 μm sections of FFPE tissues were deparaffinized with xylene and purified using solid-phase extraction. Tamoxifen and its metabolites were separated and detected by liquid chromatography–tandem mass spectrometry using multiple-reaction monitoring.ResultsThis method was linear between 0.4 and 200 ng/g for 4-hydroxy-tamoxifen and endoxifen, and 4–2,000 ng/g for tamoxifen and N-desmethyl-tamoxifen. Inter- and intra-assay precisions were <9 %, and mean accuracies ranged from 81 to 106 %. Extraction recoveries were between 83 and 88 %. The validated method was applied to FFPE tissues from two groups of patients, who received 20 mg/day of tamoxifen for >6 months, and were classified into breast tumor recurrence and non-recurrence. Our preliminary data show that levels of tamoxifen metabolites were significantly lower in patients with recurrent cancer, suggesting that inter-individual variability in tamoxifen metabolism might partly account for the development of cancer recurrence. Nevertheless, other causes such as non-compliance or stopping therapy of tamoxifen could possibly lead to the concentration differences.ConclusionsThe ability to successfully study tamoxifen metabolism in such tissue samples will rapidly increase our knowledge of how tamoxifen’s action, metabolism and tissue distribution contribute to breast cancer control. However, larger population studies are required to understand the underlying mechanism of tamoxifen metabolism for optimization of its treatment.

Derivitization, a chemical technique used to transform a compound into a similar product with a different structure required by certain detection methods, was carried out for the detection of Busulfan (Bu) in human plasma. Using a proficient scientific method and enhancing its clinical association our method furthered the pre-column derivitization technique based on isocratic high performance liquid chromatography with ultraviolet detection (HPLC-UV). In-house adjustments have been made over the course of the last two years in order to increase the efficiency and turn-around of the method assay for clinical purposes. The concentration of drug in human plasma was measured with UV detection at 278nm by employing a Waters Nov-Pak C-18 analytical column heated to 40°C. The mobile phase, run at 1.5mL/min, consisted of a mixture of 20:80 deionized water and methanol respectively. All concentration-time plasma Bu data were analyzed by a non-compartmental analysis employing linear kinetics using WinNonlin Version 4.1 software (Pharsight Corporation, Mountain View, CA, USA). Analyses were performed on 155 allogeneic Stem-Cell Transplant (SCT) patients (pts). All pts received fludarabine 50mg/m2 on days -6 to -2 and IV Bu at a “myeloablative” dose of 3.2mg/kg daily, on days -5 to -2 inclusive. Additional TBI 200cGy × 2 was given to 74 pts. Graft-vs.-host disease prophylaxis for all pts comprised cyclosporine A, “short course” methotrexate with folinic acid and Thymoglobulin (Genzyme) 4.5mg/kg administered in divided doses over 3 consecutive days pre-transplant finishing on D0. The range of area under the curve (AUC) per dose was 2184 to 7513μM·min (median 4509μM·min). Twenty nine pts (19%) had AUC < 3600μM·min and 17 (11%) had AUC > 6000μM·min. The remaining 109 (70%) were between 3600 and 6000μM·min, equivalent to an exposure of 900 – 1500μM·min per dose established as desirable from studies of qid oral Bu given with cyclophosphamide. We found that IV Bu resulted in a 3–4 fold variability in AUC with a significant number of pts with exposures deemed either too high or low. High exposures (> 6000uM*min) increase the risk of transplant-related death and there is justification for therapeutic monitoring. Our predicted adjustments obtained from a streamlined method are based on day -5 Bu doses. The adjusted doses are administered on days -3 and -2, allowing for a 24 hour turn around time between days -5 and -3.