Wide variation in po Bu absorption and metabolism within and between patients (Pts) undergoing HSCT contributes to increased relapse rate and excess toxicity. Pharmocokinetic (PK) studies in Phase 2 trials with ivBu confirmed minimal variation in area under the concentration vs time curve (AUC) in sequential doses given to individual Pts and dramatic reduction in variation in time to maximum dose, but only a modest decrease in interpatient variation. These studies suggest that dosing to achieve a precise drug exposure based upon test or first dose PK can be achieved. Attempts to accomplish this must take into consideration previously documented significant drug interactions with itraconazole and acetaminophen, both of which increase Bu levels by competing for glutathione reductase dependent metabolism, and dilantin, which produces a more slowly developing decrease in Bu level presumably due to metabolic pathway activation. We have developed a limited sampling strategy that dramatically simplifies accurate Bu PK when the iv formulation is used. We have begun studies to determine the ability of test or first dose PK parameters to accurately predict subsequent Bu levels and to determine the best dosing algorithm for ivBu use when PK expertise is not available. Pts are given the first of 16 2-hr infusions of ivBu as an outpatient 48 hr prior to 15 additional Q6h inpatient doses adjusted to achieve a target AUC on the basis of 1st dose PK. PK study is performed using a 5-sample limited sampling strategy and a single compartment, first order elimination model in WinNonlin. Pts receive lorazepam for seizure prophylaxis to avoid any dilantin effect. Of 13 Pts receiving a Bu-based auto or allo HSCT preparative regimen between April and June this year according to this schedule, 5 required no dose adjustment for dose 2, 5 required an increase and 3 required a decrease. Despite these adjustments we observed greater than 10% variation from predicted AUC for dose 2 in 5 Pts. The largest variation (30% higher than predicted) occurred in a Pt who received voriconazole between dose 1 and dose 2. The second largest variation (23% lower than predicted) occurred in a Pt who was taking an acetaminophen containing OTC product prior to dose 1, but discontinued use prior to dose 2. The third and fourth highest variations (21% and 14% higher than predicted) occurred in Pts with a rise in serum bilirubin into the abnormal range between doses 1 and 2. All of these variations were in the direction predicted by the putative cause. No other Pt had significant concomitant medication history or significant variation in serum bilirubin level between doses. The remaining Pts all had lower than predicted AUC for dose 2 (0 to -12%, median -4%). This suggests the possibility of auto-induction of metabolism. Further studies will examine whether Bu levels continue to decrease with later doses in the series by any clinically significant amount. Thus 8/9 Pts without a presumptive cause for instability in drug metabolism, including previously undocumented concomitant voriconazole, had dose 2 AUC within 10% of predicted after dose 1 PK driven adjustment. Dosing precision with ivBu is achievable, but precautions to assure absence of, or predictability of, drug interaction or changing hepatic function are necessary. Dosing precision with ivBu will improve safety and effectiveness of HSCT using Bu-based preparative regimens and greatly facilitate research into modified intensity preparative regimens for allogeneic HSCT.
Full conference title:
American Society of Hematology 45th Annual Meeting
- ASH 45th (2003)