Development of a Pharmacokinetic and Bayesian Optimal Sampling Model for Individualization of Oral Busulfan in Hematopoietic Stem Cell Transplantation

Limited Sampling Strategies for Estimating Busulfan Area Under the Concentration-Time Curve: Based on Peak and Trough Concentrations in Saliva

Therapeutic drug monitoring for busulfan is currently performed by multiple plasma sampling. Saliva is considered a noninvasive therapeutic drug monitoring matrix. This study aimed to investigate intravenous busulfan pharmacokinetics (PK) in plasma and saliva, and establish a limited sampling strategy (LSS) for predicting the area under the concentration-time curve from time zero to infinity in plasma (AUC0-∞,p) by using saliva samples. Therefore, the PK of busulfan was studied in 37 Chinese patients. Pearson correlation analysis was used to evaluate the correlation between the AUC of busulfan in plasma and saliva. LSS models were established by the multiple linear regression analysis. The prediction error, the mean prediction error, and the root mean square error were used to evaluate the predictive accuracy. The agreement between the predicted and observed AUC0-∞ in saliva was investigated by the intraclass correlation coefficient and Bland-Altman analysis. The accuracy and robustness of the models were evaluated by using the bootstrap procedure. The result of PK analysis 62.2% of patients (23/37) was within the target range of AUC0-∞,p . A good correlation between saliva and plasma busulfan AUC0-∞ was observed (r = 0.63, p < .01). The bias and precision of the models 7 and 13 were less than 15%. The intraclass correlation coefficient exceeded 0.9, and the limits of agreement were within ±15%. The 2-point LSS model in saliva is a convenient and desirable approach to predict the AUC0-∞ of 4 times daily intravenous busulfan in plasma, which can be used to design personalized dosing for busulfan.

Utilization of a Population Pharmacokinetic Model to Improve a Busulfan Test-Dose Strategy in Allogeneic Hematopoietic Cell Transplant Recipients

Busulfan is an alkylating agent used as part of conditioning chemotherapy regimens prior to allogeneic hematopoietic cell transplant (allo-HCT). Pharmacokinetic (PK)-guided test-dose strategies have been shown to improve the number of patients achieving busulfan exposure goals and improve clinical outcomes. However, current practices require extensive PK sampling. In this study, PK data were retrospectively collected from busulfan drug monitoring records from adult allo-HCT recipients who received once-daily intravenous busulfan at the University of North Carolina Medical Center (UNCMC). A population pharmacokinetic (popPK) model was developed to identify sources of interindividual variability and evaluate alternative PK sampling strategies. A 2-compartment model, with covariate effects of actual body weight and sex, best described the data. The typical value of clearance for an 83 kg male was estimated to be 11.21 L/h. Fifty-nine percent of allo-HCT recipients were estimated to have met the UNCMC institutional myeloablative conditioning (MAC) exposure goal based on model post hoc estimates of clearance using all PK samples obtained following MAC dosing. Fifty-seven percent of patients were estimated to have met this goal based on post hoc estimates using a single PK sample. Our results indicate once-daily, intravenous busulfan PK in adult allo-HCT recipients receiving MAC dosing can be reasonably described by a popPK model, and the use of a sparse PK sampling strategy may be feasible for determining target exposure attainment following MAC dosing. Use of a popPK model and sparse PK sampling strategy to carry out busulfan test-dose procedures could reduce health care costs and inconvenience to patients.

Predicting busulfan exposure in patients undergoing hematopoietic stem cell transplantation using machine learning techniques

PURPOSE

This study aimed to establish an optimal model to predict the busulfan (BU) area under the curve at steady state (AUCss) by using machine learning (ML).

PATIENTS AND METHODS

Seventy-nine adult patients (age ≥18 years) who received BU intravenously and underwent therapeutic drug monitoring from 2013 to 2021 at Fujian Medical University Union Hospital were enrolled in this retrospective study. The whole dataset was divided into a training group and test group at the ratio of 8:2. BU AUCss were considered as the target variable. Nine different ML algorithms and one population pharmacokinetic (pop PK) model were developed and validated, and their predictive performance was compared.

RESULTS

All ML models were superior to the pop PK model (R2 = 0.751, MSE = 0.722, 14 and RMSE = 0.830) in model fitting and had better predictive accuracy. The ML model of BU AUCss established through support vector regression (SVR) and gradient boosted regression trees (GBRT) had the best predictive ability (R2 = 0.953 and 0.953, MSE = 0.323 and 0.326, and RMSE = 0.423 and 0.425).

CONCLUSION

All the ML models can potentially be used to estimate BU AUCss with the aim of facilitating rational use of BU on the individualized level, especially models built by SVR and GBRT algorithms.

Evaluation of a Nanoparticle-Based Busulfan Immunoassay for Rapid Analysis on Routine Clinical Analyzers

BACKGROUND

Busulfan is an alkylating agent used in allogeneic hematopoietic stem cell transplantation for various malignant and nonmalignant disorders. Therapeutic drug monitoring of busulfan is common because busulfan exposure has been linked to veno-occlusive disease, disease relapse, and failed engraftment. The authors developed an automated immunoassay, along with stable calibrators and controls, and quantified busulfan in sodium heparin plasma.

METHODS

The authors evaluated a homogenous nanoparticle immunoassay, the MyCare Oncology Busulfan Assay Kit (Saladax Biomedical, Inc), for precision, sensitivity, accuracy, and linearity on an open channel clinical chemistry analyzer; they compared the method with 2 mass spectrometry methods (liquid chromatography-tandem mass spectrometry and gas chromatography/mass spectrometry), using anonymized, remnant patient samples.

RESULTS

The coefficients of variation for repeatability and within-laboratory precision were ≤9.0%. The linear range was 150-2000 ng/mL; samples up to 6000 ng/mL can be measured with sample dilution. Measured values deviated by ≤14% from assigned values. Comparison between validated mass spectrometry methods resulted in a correlation coefficient R ≥ 0.995.

CONCLUSIONS

The MyCare Busulfan Assay Kit shows the precision, accuracy, linearity, and test range for performing busulfan concentration measurements in sodium heparin plasma on routine clinical chemistry analyzers.

Limited Sampling Strategies Supporting Individualized Dose Adjustment of Intravenous Busulfan in Children and Young Adults

BACKGROUND

Therapeutic drug monitoring (TDM) for busulfan supports dose adjustment during conditioning for stem cell transplantation. The authors aimed to develop and validate limited sampling strategies (LSS) of 4-5 samples for a precise estimation of the area under concentration (AUC)-time curve of busulfan, in plasma as an alternative to an intensive sampling strategy (ISS) requiring 9-10 samples.

METHODS

ISS TDM data from 297 patients (≤18 years of age) were used. AUCLSS was calculated using the trapezoidal rule and multiple linear regression (MLR). Unlike more complex modeling methods, MLR does not require sophisticated software or advanced training of personnel. MLR coefficients were estimated in the development subset containing randomly selected 50% of the records and were then used to calculate the AUCLSS of the remaining records (the validation subset). The agreement between dose adjustment recommendations (DAR) based on ISS and LSS, in the validation subset, was evaluated by a Bland-Altman analysis. A DAR deviating from an ISS-based reference by <15% was deemed acceptable.

RESULTS

Twelve LSSs were acceptable. Sampling at 0, 120, 180, and 240 minutes after the start of the second infusion (LSS15) yielded the best performance, with DAR deviating from the reference by <10% for 95% of cases; the AUCLSS was determined as follows: AUCLSS = 74.7954 × C(0) + 81.8948 × C(120) + 38.1771 × C(180) + 138.1404 × C(240) + 54.1837. This LSS and LSS13 performed similarly well in an independent external validation.

CONCLUSIONS

MLR-based estimates of AUCLSS provide DARs that deviate minimally from the reference. LSSs allow the reduction of patient discomfort, a ∼50% reduction of TDM-related workload for nursing staff and blood loss and a ∼25% reduction in laboratory workload. These benefits may encourage wider use of busulfan TDM, supporting safe and efficacious personalized dosing.

Impact of GSTA1 Polymorphisms on Busulfan Oral Clearance in Adult Patients Undergoing Hematopoietic Stem Cell Transplantation

Background: Busulfan pharmacokinetics exhibit large inter-subject variability. Our objective was to evaluate the influence of glutathione S-transferase A1 (GSTA1) gene variants on busulfan oral clearance (CLo) in a population of patients undergoing hematopoietic stem cell transplantation. Methods: This is a quasi-experimental retrospective study in adult patients (n = 87 included in the final analyses) receiving oral busulfan. Pharmacokinetics data (area under the plasma concentration-time curve (AUC) determined from 10 blood samples) were retrieved from patients’ files and GSTA1 *A and *B allele polymorphisms determined from banked DNA samples. Three different limited sampling methods (LSM) using four blood samples were also compared. Results: Carriers of GSTA1*B exhibited lower busulfan CLo than patients with an *A/*A genotype (p < 0.002): Busulfan CLo was 166 ± 31, 187 ± 37 vs. 207 ± 47 mL/min for GSTA1*B/*B,*A/*B and *A/*A genotypes, respectively. Similar results were obtained with the tested LSMs. Using the standard AUC method, distribution of patients above the therapeutic range after the first dose was 29% for GSTA1*A/*A, 50% for *A/*B, and 65% for *B/*B. The LSMs correctly identified ≥91% of patients with an AUC above the therapeutic range. The misclassified patients had a mean difference less than 5% in their AUCs. Conclusion: Patients carrying GSTA1 loss of function *B allele were at increased risk of overdosing on their initial busulfan oral dose. Genetic polymorphisms associated with GSTA1 explain a significant part of busulfan CLo variability which could be captured by LSM strategies.

Accurate Prediction of Initial Busulfan Exposure Using a Test Dose With 2- and 6-Hour Blood Sampling in Adult Patients Receiving a Twice-Daily Intravenous Busulfan-Based Conditioning Regimen

This study aimed to predict the area under the curve (AUC) of the initial busulfan dose using a test dose with the sparse sampling scheme in adult patients who underwent hematopoietic cell transplant. A test dose of 0.8 mg/kg busulfan was used 2 days before twice-daily intravenous busulfan-based conditioning regimens were administered. The AUC and the clearance (CL) were calculated for both the test dose and the first dose (AUC_T , CL_T , AUC_1, and CL₁ ) by noncompartmental analysis. The sparse sampling schemes of the test dose were developed by Bayesian method based on the population pharmacokinetic model. The optimal sparse sampling schemes were determined by evaluating the mean prediction error, the root mean square error, the absolute mean prediction error, and Bland-Altman plot. The mean AUC₁ was 7.20 ± 1.48 mg • h/L, which ranged from 4.70 to 9.46 mg • h/L. The AUC₁ was below the therapeutic concentration of 7.38 mg • h/L in 45% (9 of 20) of the patients. The CL_T of 3.05 ± 0.56 mL/min/kg was not significantly different with the CL₁ of 3.03 ± 0.69 mL/min/kg (P = .901). A sampling scheme at 2 and 6 hours after the test dose was developed to predict the AUC_T (mean prediction error of 1.64%, root mean square error of 6.17%, and absolute mean prediction error of 4.94%). Additionally, the Bland-Altman plot showed that the 2-sampling scheme provided an acceptably accurate prediction of the AUC₁ . A test dose with a 2-sampling scheme was sufficient to personalize the initial busulfan dosing in hematopoietic cell transplant recipients.

Evaluation of Limited Sampling Methods for Oral Busulfan Pharmacokinetic Monitoring in Adult Patients Undergoing Hematopoietic Stem Cell Transplantation

BACKGROUND

Monitoring busulfan area under the plasma concentration-time curve (AUC) to establish the dose regimen for stem cell transplantation desirable to achieve efficacy while avoiding toxicity.

OBJECTIVE

Our objective was to compare AUCs calculated by 18 limited sampling methods (LSMs) from 2 to 5 samples to reference AUCs determined from 10 samples in a retrospective study of 103 adult patients receiving oral busulfan. LSMs using 2 or 3 samples were ineffective.

METHODS

Four LSMs using 4 or 5 blood samples that accurately characterized busulfan AUC were identified. The best 2 methods were obtained with sampling at 0.5, 1, 2, 4, and 6 hours and after 1, 1.5, 2, 4, and 6 hours postdose. For these LSMs, the incidence of 20% difference between AUCs from LSMs and reference AUCs was less than 1.3%.

CONCLUSIONS

Effective and safe determination of AUC for oral busulfan can be made with strategies using only 4 or 5 concentration timepoints.

Busulfan dosing algorithm and sampling strategy in stem cell transplantation patients

AIM

The aim of this investigation was to develop a model-based dosing algorithm for busulfan and identify an optimal sampling scheme for use in routine clinical practice.

METHODS

Clinical data from an ongoing study (n = 29) in stem cell transplantation patients were used for the purposes our analysis. A one compartment model was selected as basis for sampling optimization and subsequent evaluation of a suitable dosing algorithm. Internal and external model validation procedures were performed prior to the optimization steps using ED-optimality criteria. Using systemic exposure as parameter of interest, dosing algorithms were considered for individual patients with the scope of minimizing the deviation from target range as determined by AUC(0,6 h).

RESULTS

Busulfan exposure after oral administration was best predicted after the inclusion of adjusted ideal body weight and alanine transferase as covariates on clearance. Population parameter estimates were 3.98 h(-1), 48.8 l and 12.3 l h(-1) for the absorption rate constant, volume of distribution and oral clearance, respectively. Inter-occasion variability was used to describe the differences between test dose and treatment. Based on simulation scenarios, a dosing algorithm was identified, which ensures target exposure values are attained after a test dose. Moreover, our findings show that a sparse sampling scheme with five samples per patient is sufficient to characterize the pharmacokinetics of busulfan in individual patients.

CONCLUSION

The use of the proposed dosing algorithm in conjunction with a sparse sampling scheme may contribute to considerable improvement in the safety and efficacy profile of patients undergoing treatment for stem cell transplantation.

Optimal sampling strategy development methodology using maximum a posteriori Bayesian estimation

Maximum a posteriori Bayesian (MAPB) pharmacokinetic parameter estimation is an accurate and flexible method of estimating individual pharmacokinetic parameters using individual blood concentrations and prior information. In the past decade, many studies have developed optimal sampling strategies to estimate pharmacokinetic parameters as accurately as possible using either multiple regression analysis or MAPB estimation. This has been done for many drugs, especially immunosuppressants and anticancer agents. Methods of development for optimal sampling strategies (OSS) are diverse and heterogeneous. This review provides a comprehensive overview of OSS development methodology using MAPB pharmacokinetic parameter estimation, determines the transferability of published OSSs, and compares sampling strategies determined by MAPB estimation and multiple regression analysis. OSS development has the following components: 1) prior distributions; 2) reference value determination; 3) optimal sampling time identification; and 4) validation of the OSS. Published OSSs often lack all data necessary for the OSS to be clinically transferable. MAPB estimation is similar to multiple regression analysis in terms of predictive performance but superior in flexibility.

Therapeutic drug monitoring in cancer chemotherapy

For a select number of drugs, proper management of patients includes monitoring serum or plasma concentrations of the drugs and adjusting the doses accordingly - this practice is referred to as therapeutic drug monitoring (TDM). The need for TDM arises when pharmacokinetic variability of drugs is not easily accounted for by common clinical parameters. Many chemotherapeutic drugs have large interindividual variability, yet TDM is not commonplace in chemotherapy management. This review will discuss pharmacokinetics in the context of chemotherapeutic drugs, examine the few instances where TDM is currently used in the field of oncology and propose other drugs where TDM might be useful for dose adjustments in the management of chemotherapy.

A pilot study of reduced toxicity conditioning with BU, fludarabine and alemtuzumab before the allogeneic hematopoietic SCT in children and adolescents

We report the results of a pilot study of a BU-fludarabine-alemtuzumab (BFA)-reduced toxicity conditioning (RTC) followed by allogeneic hematopoietic SCT (AlloHSCT) in 12 children and adolescents (<21 years) with malignant and non-malignant diseases. Stem cell sources were: two unrelated cord blood, one unrelated BM, two related and seven unrelated PBSC. Positive CD34 selection was performed in five unrelated PBSC grafts. RCT was carried out with BFA, and GVHD prophylaxis was FK506 and mycophenolate mofetil. The median time for neutrophil and platelet engraftment was 16 and 31 days, respectively. The P of developing ≥ grade II, ≥ grade III aGVHD and cGVHD was 41.6, 25 and 9%, respectively. Only 1 out of 12 developed ≥ grade III toxicity. There was one primary and no secondary graft failure. Mixed donor chimerism on day 100 and 1 year was median 99 and 96%, respectively; ≥ 90% of recipients achieved ≥ 80% donor chimerism. The 3-year overall survival (OS) in all patients was 91.7 ± 8% (100% for malignant vs. 80% for non-malignant diseases, ns). In all, 11 (91%) patients remain alive at median 2.8 (0.3-6.8) years. RTC followed by AlloHSCT, based on BFA conditioning, is feasible and tolerable in children and adolescents, and results in prompt achievement of durable mixed donor chimerism and excellent OS.

Development of a population pharmacokinetics-based sampling schedule to target daily intravenous busulfan for outpatient clinic administration

Therapeutic drug monitoring of daily intravenous (IV) busulfan currently requires hospital admission. Population pharmacokinetic modeling and determination of an optimal pharmacokinetic sampling schedule over 6 hours could allow for personalizing these busulfan doses in the outpatient clinic. A retrospective evaluation of daily IV busulfan pharmacokinetics was conducted in 37 adults. SPK and NONMEM software were used to estimate the population pharmacokinetic parameters. Subsequent to model building, the area under the concentration-time curve (AUC) was computed using NONMEM. A 1-compartment model best fit the data. The optimal 6-hour outpatient sampling schedule was constructed using a simulation approach that sought to minimize scaled mean squared error for the clearance and volume parameters for each simulated individual. The best sampling times were 2.75, 3, 3.25, 5.5, 5.75, and 6 hours from the start of a 3-hour infusion. With these sampling times, the maximum a posteriori (MAP) Bayesian estimation was superior to maximum likelihood estimation with more samples. An individual patient's busulfan AUC and pharmacokinetic parameters may be accurately estimated with an outpatient sampling schedule that is used in conjunction with MAP Bayesian estimation, with a parameter prior based on population pharmacokinetic modeling. Prospective validation of this approach is needed.

Busulfan in hematopoietic stem cell transplant setting

This paper focuses primarily on the data published in the last decade about the pharmacokinetics and pharmacodynamics of oral and intravenous (i.v.) busulfan, therapeutic drug monitoring and clinical outcome in hematopoietic stem cell transplant (HCT) patients. Busulfan is commonly used in HCT as it is toxic to the marrow. Busulfan is available as oral or i.v. formulation. The most common significant toxicity of busulfan is sinusoidal obstruction syndrome. Even with the introduction of i.v. busulfan, variability in the systemic concentrations of busulfan after weight-based dosing and the association between busulfan plasma exposure and outcome in HCT patients have led to the continued use of therapeutic drug monitoring of busulfan. New strategies for personalizing busulfan dosing are being studied to maximize the use of busulfan for optimal disease control with the least toxicity to HCT patients. One such strategy currently being evaluated is if busulfan clearance can be accurately predicted by genetic polymorphism of glutathione S-transferase (GST), with the currently available data suggesting that GST polymorphisms cannot be used to personalize busulfan dosing.

Influence of Underlying Disease on Busulfan Disposition in Pediatric Bone Marrow Transplant Recipients: A Nonparametric Population Pharmacokinetic Study

Busulfan is an alkylating agent used in a conditioning regimen prior to bone marrow transplantation. Busulfan has a narrow therapeutic index, giving rise to major liver toxicity (veno-occlusive disease), and a wide interpatient and intrapatient pharmacokinetic variability. This report presents the results of a population pharmacokinetic analysis leading to models based on underlying diseases requiring bone marrow transplantation. One hundred children received oral busulfan-based conditioning regimens between March 1998 and February 2006. Busulfan pharmacokinetic parameter estimates (Ka, first order absorption rate constant; Vs, volume of distribution related to the body weight; and Cl/F, apparent clearance) were estimated by using the nonparametric adaptative grid (NPAG) algorithm in patients divided into four groups according to initial diagnosis: metabolic diseases, hemoglobinopathies, hematological malignancies, and immune deficiencies. Ka and Vs did no differ significantly in the four subgroups. Cl/F and areas under the plasma concentration curve were significantly different in the four groups. Cl/F was significantly higher in the hemoglobinopathies group (P = 0.002), with a mean value of 7.78 L . h, whereas the immune deficiencies group was characterized by the lowest Cl/F (3.59 L . h). Interindividual variability was shown by high interindividual parameter percent coefficients of variation (CV%) but, nevertheless, with less diversity in the population parameter distributions for Vs in the three subgroups-metabolic diseases, hemoglobinopathies, and malignant diseases-and in Cl/F for patients with hemoglobinopathies. The fit was good for busulfan concentration predictions based on Bayesian individual posterior values, with little bias and good precision. In comparison with the overall population, the only model of subgroup presenting a greater precision was patients with hemoglobinopathies (P = 0.002). Use of these more specific models of a given disease may well result in more accurate individualization of busulfan dose regimens, especially in very sparse blood sampling situations.