Pharmacokinetics/Pharmacodynamic Relationship in Busulfan Conditioning Regimen: Results from a Large Pediatric Cohort Undergoing Hematopoietic Stem-Cell Transplantation

Background: Busulfan (Bu) is the corner stone of hematopoietic stem-cell transplantation (HSCT) regimens with a narrow therapeutic window (TW). Graft rejection or toxicities are reported according to plasmatic exposure. In very young children, Bu exhibits large pharmacokinetic (PK) variability. Bu clearance was demonstrated to be non-linearly related to body weight (BW) and thus BW is used to optimally define the Bu dosage in children [Vassal et al., CCP 2006]. Search for additional data to confirm the TW and to describe clinical outcomes is still a topical question to better understand Pk/Pd relationship and to safely use Bu in young children and infants.

Objective: To study the Pk/Pd relationship of Bu in pediatric recipients of bone marrow transplantation (BMT) receiving Bu-based conditioning regimens (CR). To correlate early toxicities (mainly hepatic veno-occlusive disease (VOD), non infectious pulmonary disease (niPD) and outcome i.e. overall survival (OS) at last follow up with type of CR, the underlying diseases, age at transplantation and Pk of busulfan

Patients and Method: This multicenter prospective observational study included 307 pts transplanted between 2006 and 2013 from 14 French Pediatric BMT units; median age at transplantation was 18.4 months [1.3-289], with a median BW of 11.3 kg [3.4-82]. 100 pts were younger than 1 year, 71 pts < 9 kg and 171 pts < 16 kg. Patients were mostly affected by non-malignant diseases (primary immune deficiency (n=143), inherited metabolic disorders (n=50), hemoglobinopathies (n=20) while 78 patients suffered from malignant disease. 257 patients received allogenic HSCT (genoidentical donor n=79, matched unrelated donor n=33, mismatch unrelated donor n=18, haploidentical intra-familial donor n=59, other mismatched intra familial donor n=8 or unrelated cord blood n=51) and 41 autologous BMT. All patients received Bu-based conditioning regimen in combination with cyclophosphamide (BuCy n=119), fludarabine (BuFlu n=88), melphalan (BuMel n=36) or thiotepa (BuTTP n=3). 40 patients received BuFlu associated with a second alkylating agent (Mel, TTP or Cy), 12 patients received BuCy associated with a third agent (Mel or VP16). Serotherapy was given in 70% of the patients. Starting doses of Bu were given according to the European SmPC or EBMT-ESID recommendations. The median posology was 1 mg/kg 4x/day for 4 days [0.6-1.3 mg/kg].

PK was assessed on plasma samples with area under the curve (AUC) evaluation from the 1^st(D1), 9^th (D9) and 13^th (D13) dose in 150 patients, D1 and D9 in 40 patients, D1 and D13 in 46 patients while 62 patients were monitored on D1 only (684 PK datasets). PK analysis was performed using Non linear Mixed Effect Modelling. A one-compartment PK model suitably fitted the concentrations vs time data. Median follow up is 27 months (0.33 to 96 months post BMT).

Results: At D1, 67% of patients were within the therapeutic window (TW) [900-1500 µM.min] and 66% of patients reached the cumulative TW [14400-24000 µM.min]. Incidence of VOD and niPD were respectively 35% and 14%. Both toxicities correlated with type of CR and age <1 year at transplantation. Incidence of aGVH was 33%. However, these incidences were very different according to the CR. BuFlu showed the lowest incidence compared to other combinations of drugs (p < 0.001 and =0.022 for VOD and niPD, respectively). When using a second alkylating agent, incidence of VOD reached 41% (vs 19% in BuFlu, p<0.001) and niPD 17% (vs 8% in BuFlu, p<0.040). Toxicities were also related to BW, with 48% vs 30% (p<0.013) for VOD below or above 9 kg, and respectively, 28% vs 9% (p<0.0002) for niPD. This tendency was also found with a 15 kg cut-off for aGVH. No correlation was found with Bu AUC. However, there was a trend for lower OS (HR 1.60 [1.00-2.58], p=0.051) in patients with AUC D1 <900 µM.min compared to patients with AUC D1 > 900 µM.min. OS was 69.1% at last follow up and was not significantly associated with CR and age at transplantation.

Conclusion: In this large series of pediatric BMT recipients, we found that toxicities as VOD and niPD correlated with type of CR and age at transplantation in univariate analysis. Combination of 2 alkylating agents with Bu or BuFlu and transplantation < 1 year of age were associated with the highest incidence of toxicites. The TW targeting performance was improved in this cohort compared to previous reported data (Paci et al. TDM 2012). Specific TW for each group of pathologies will be proposed.