Pharmacometabolomics for Predicting Variable Busulfan Exposure in Pediatric HSCT Patients

Optimal dosing for busulfan is important for minimization of systemic toxicity from overexposure and graft failure or relapse from underexposure. Herein, we investigated potential markers for predicting individual variation in the pharmacokinetics of busulfan, and suggested possible mechanism for inter-individual variability by using pharmacometabolomics.

Fifty-nine pediatric patients undergoing busulfan-based conditioning chemotherapy for hematopoietic stem cell transplantation were divided into three groups according to the area under the concentration-time curve (AUC) of busulfan; low-, medium-, and high-AUC group. Nontargeted metabolic profiling of baseline urine samples showed that deferoxamine metabolites were abundant, while 2 acylcarnitines and phenylacetylglutamine were significantly lower in high-AUC group, compared with low-AUC group. Higher level of deferoxamine, an iron-chelating agent for the patients with a high serum ferritin level during the conditioning chemotherapy, suggested pharmacokinetic interaction between serum ferritin level and busulfan exposure. Retrospective analysis of the correlation between serum ferritin level and busulfan AUC showed positive correlation in 130 pediatric patients. The optimal busulfan dose to meet the target AUC of 18,750 μg*h/L/day was calculated to be 119.7 ± 30.1 mg/m² in patients with ferritin < 1,000 ng/mL and 106.1 ± 29.3 mg/m²in patients with ferritin ≥ 1,000 ng/mL (P=0.021).

Mechanismly, previous studies have indicated that ferritin, acylcarnitine and phenylacetylglutamine are closely associated with liver function. Increased serum ferritin is thought to be responsible for an increased production of oxygen free radicals and activation of GSH turnover, which means reduction of GSH levels in both plasma and erythrocytes and this depletion seems to be related to the decrease of busulfan metabolism. Down regulation of acylcarnitines is associated with deregulation of mitochondrial fatty acid β-oxidation by hepatic injury. PAGN, a marker of waste nitrogen scavenger, was down-regulated which indicates ammonia-related metabolism could also be involved in busulfan exposure. Hyperammonemia, a clinical condition of elevated ammonia levels, is mainly lead by liver cell damage.

Taken together, our findings demonstrate that elevated serum ferritin levels are a potential biomarker correlated with busulfan exposure and provide some evidence that busulfan metabolism seems to decrease in the patients with reduced liver function.