Pharmacokinetics of Cyclophosphamide Metabolites Influence Outcome in Patients with β-Thalassemia Major Undergoing Allogeneic HSCT

Abstract 1941

Busulfan in combination with cyclophosphamide (Cy) is the commonly used conditioning regimen for hematopoietic stem cell transplantation (HSCT) for various hematological diseases. Cy, a prodrug, undergoes hepatic biotransformation to 4-hydroxy cyclophosphamide (4-HCy) and subsequently to its active metabolite, phosphoramide mustard (PM) and carboxyethyl PM (CEPM), a nontoxic oxidation product of HCY/aldophosphamide. Though toxic complications like hemorrhagic cystitis (HC) and hepatic sinusoidal obstruction syndrome (SOS) have been associated with metabolites of Cy such as acrolein and CEPM, respectively, there is no data correlating pharmacokinetics (PK) of Cy, HCy or CEPM with toxicity and outcome of HSCT for ß thalassemia major. Aim of the present study was to evaluate Cy, 4-HCy and CEPM PK and the influence of these PK parameters on clinical outcome in patients with ß thalassemia major undergoing HSCT. Between January 2001 to June 2009, out of the 168 HSCT for thalassemia major conditioned with Bu/Cy regimen (including 8 second transplants), 90 patients for whom PK samples were collected were included in this study. Cy was administered at 50 mg/kg/day for 4 days (days −5 to −2) following 4 days of busulfan (days −9 to −6). Peripheral blood samples were collected during Cy infusion at various time interval and plasma samples were stored for Cy, HCy and CEPM PK analysis. Levels of Cy and 4-HCy were measured by high performance liquid chromatography, and CEPM, by a modified LCMSMS method. The population PK estimates were determined using non-linear mixed effects modeling analysis performed with Monolix (version 3.1, www.monolix.org). Specifically, a compartmental model which included two compartments each for Cy, HCy and CEPM was used to describe the data. Clinical outcome endpoints including graft rejection, event free survival (EFS), overall survival (OS), transplant related mortality (TRM), SOS, and HC were evaluated using standard criteria. The influence of Cy and metabolite PK on clinical outcome endpoints were compared using logistic regression analysis.

Age range of the patients was 2 to 24 years. Four patients belonged to Lucarelli risk class I, 49 class II and 37 class III. Based on risk stratification that we have defined using liver size and age (high risk: age >7 years and liver size >5cm; and the rest as low risk; Mathews et al, 2007), 39 were in low risk, 40 intermediate and 11 were high risk. Overall incidence of OS, EFS, rejection, TRM, SOS, HC in this cohort was 77, 70, 14, 10, 18 and 31% respectively. It should be noted that this does not completely represent the outcome of HSCT in thalassemia during this period, as only patients with available Cy PK analysis were included for analysis. The high risk patients had significantly reduced OS (RR 2.59; p=0.04), EFS (RR 2.23; p=0.058), increased risk of TRM (RR: 3.56; p=0.03) and HC (RR: 3.19; p=0.036) compared to others, while this was not significantly different with respect to Lucarelli class except for increased incidence of HC in Lucarelli class III patients (RR: 2.6; p=0.04).

Upon univariate analysis, there was significantly increased Cy AUC (1887 ng*h/ml; range: 679–8546; vs. 1544, range: 662–4434; p= 0.028) in patients who developed HC compared to those who did not. There was significantly decreased HCy AUC (median 5.172 ug*L/h, range: 0.795–6.457; vs. 6.224; 2.536–12.003 p=0.007) in patients who died compared to those who are alive; similar but more significant association was seen with EFS as well. There was decreased CEPM Cl/F in those who rejected the grafts (0.013 vs. 0.028 L/h/Kg; p=0.016), while it was significantly increased in patients who developed SOS (median 0.029; vs. 0.013, range: L/h/Kg; p=0.05). Upon forward stepwise multivariate analysis including all the Cy and metabolite PK parameters, only HCy AUC significantly influenced EFS and OS in these patients. We show here for the first time that Cy and metabolite PK influences HSCT outcome in a uniform cohort of patients with thalassemia major. However, due to the complex metabolic pattern of Cy, and the association of metabolite PK instead of the levels of the parent compound with outcome, the possibility of targeted dose adjustment of Cy to improve HSCT outcome may be more challenging than targeted dose adjustment of other drugs used in HSCT.