Intravenous Busulfan, Dimethylacetamide and neurotoxicity after high-dose pretransplant conditioning chemotherapy

INTRODUCTION Pursuing more efficacious and less toxic allogeneic stem-cell transplant (allo-SCT) conditioning therapy, intravenous (IV) Busulfan, BusulfexTM, (Bu) was introduced in 1999 [1]. Its combinations with Cyclophosphamide (Cy) or with the nucleoside analog Fludarabine (Flu), have led for more than a decade high-dose chemotherapy for allo-SCT for patients with acute and chronic myeloid leukemia (AML, CML) and myelodysplastic syndrome (MDS) [2–4]. It is increasingly being advocated in high dose chemotherapy (HDC) with autologous SCT for relapsed Hodgkin and non-Hodgkin lymphomas and myeloma [5–8], where it may confer an advantage over previous standard of care with BCNU, etoposide, cytarabine and melphalan (BEAM) and high-dose Melphalan, respectively [6, 7]. Regarding allo-SCT for AML and MDS, Bu-based conditioning results in better overall survival than Cy-total body irradiation (TBI) [9, 10], and yields lower risk for severe graft versus host disease [11]. While IV Bu represents an improvement over TBI, there remain concerns with the formulation, initially expressed by the FDA expert reviewer in 1999 [1]. These concerns were not related to busulfan itself, but rather to its solvent vehicle including N,N-dimethylacetamide (DMA) [12, 13]. In contrast, the sponsor, Orphan Medical Inc. (OMI) (Minnetonka, MN), claimed a DMA-based solvent to be safe, based on a published trial using DMA as an anti-cancer agent [14]. However, in his review of the phase II data underlying the approval of Busulfex, the FDA expert examiner surmised that the total amount of DMA given in HDC may be enough to cause significant toxicities [1]. Occupational DMA exposure was associated with both neurological and hepatic toxicity, and in experimental systems it also caused reproductive toxicity [14–16]. In reference to possible DMA toxicity, the phase II study report submitted by OMI listed a high incidence of neurological toxicities: 84% insomnia, 72% anxiety, 30% dizziness and 23% depression following IV Bu infusion. Fewer patients suffered more serious problems, including delirium (2%), agitation (2%), confusion (11%), hallucinations (5%) and encephalopathy (2%) [1]. In our use of IV Bu-based conditioning, we recorded six patients who developed hallucinations, agitation, delirium, and in two cases deep stupor. These neurologically deranged states lasted for almost 2 months in one patient before he recovered, while in one patient it deteriorated into an encephalopathy with progressive, irreversible coma before he expired 5 weeks post-SCT. Two patients recovered within 10 days from the onset of the problems; in two patients altered mental status started days after recovery from treated bacteremia (Table 1).


INTRODUCTION
Pursuing more efficacious and less toxic allogeneic stem-cell transplant (allo-SCT) conditioning therapy, intravenous (IV) Busulfan, Busulfex™, (Bu) was introduced in 1999 [1]. Its combinations with Cyclophosphamide (Cy) or with the nucleoside analog Fludarabine (Flu), have led for more than a decade high-dose chemotherapy for allo-SCT for patients with acute and chronic myeloid leukemia (AML, CML) and myelodysplastic syndrome (MDS) [2][3][4]. It is increasingly being advocated in high dose chemotherapy (HDC) with autologous SCT for relapsed Hodgkin and non-Hodgkin lymphomas and myeloma [5][6][7][8], where it may confer an advantage over previous standard of care with BCNU, etoposide, cytarabine and melphalan (BEAM) and high-dose Melphalan, respectively [6,7]. Regarding allo-SCT for AML and MDS, Bu-based conditioning results in better overall survival than Cy-total body irradiation (TBI) [9,10], and yields lower risk for severe graft versus host disease [11]. While IV Bu represents an improvement over TBI, there remain concerns with the formulation, initially expressed by the FDA expert reviewer in 1999 [1]. These concerns were not related to busulfan itself, but rather to its solvent vehicle including N,N-dimethylacetamide (DMA) [12,13]. In contrast, the sponsor, Orphan Medical Inc. (OMI) (Minnetonka, MN), claimed a DMA-based solvent to be safe, based on a published trial using DMA as an anti-cancer agent [14]. However, in his review of the phase II data underlying the approval of Busulfex, the FDA expert examiner surmised that the total amount of DMA given in HDC may be enough to cause significant toxicities [1]. Occupational DMA exposure was associated with both neurological and hepatic toxicity, and in experimental systems it also caused reproductive toxicity [14][15][16].
In reference to possible DMA toxicity, the phase II study report submitted by OMI listed a high incidence of neurological toxicities: 84% insomnia, 72% anxiety, 30% dizziness and 23% depression following IV Bu infusion. Fewer patients suffered more serious problems, including delirium (2%), agitation (2%), confusion (11%), hallucinations (5%) and encephalopathy (2%) [1]. In our use of IV Bu-based conditioning, we recorded six patients who developed hallucinations, agitation, delirium, and in two cases deep stupor. These neurologically deranged states lasted for almost 2 months in one patient before he recovered, while in one patient it deteriorated into an encephalopathy with progressive, irreversible coma before he expired 5 weeks post-SCT. Two patients recovered within 10 days from the onset of the problems; in two patients altered mental status started days after recovery from treated bacteremia ( Table 1).

DISCUSSION
The current Bu-formulation with DMA-PEG is safer and better tolerated than oral Bu in HDC [9][10][11], but there are lingering concerns that may be related to DMA [1,13]. Unfortunately, it is impossible to evaluate IV Bu with and without DMA, because the only currently approved IV Bu formulation contains DMA. We made the following observations when we changed from oral to IV Bu in HDC: First, there appears not only to be a reduction in venoocclusive disease/sinusoidal obstruction syndrome (VOD/ SOS) of the liver, but also a change in its clinical picture. Classical VOD/SOS has a rapid onset after SCT, typically within the first 3-4 weeks, followed by a gradual decline of the patient's condition, leading to hepato-renal/multiorgan failure and death within a few weeks [17][18][19], whereas milder cases had a subacute onset followed by gradual recovery. This has commonly been replaced by an episode of "silent" transaminitis and hyperbilirubinemia occurring within 10-14 days after IV Bu in 30-40% of the patients, which typically resolves within 7-10 days. Fewer cases of serious treatment-related VOD/SOS manifest themselves either early (as in classical VOD) or later, with some patients developing liver failure as late as 8-12 weeks after SCT with subsequent decline and death [4,[20][21][22][23][24]. We hypothesize that this delayed presentation of serious, life-threatening VOD/SOS may be due to an interaction between DMA and Bu in the generation of liver toxicity, as oral Bu-(Cy) associated VOD historically manifested itself early, within the first few weeks after exposure. Delayed onset of serious liver toxicity, 8-12 weeks after exposure, is reminiscent of that described by Choi after occupational DMA exposure [15].
Second, there is concern about potentially serious neurological toxicity with the DMA-PEG formulation. In the pivotal phase II studies, OMI reported that the maximum DMA dose given as part of IV Bu, Busulfex™, was about 42% of the maximum tolerated dose on a mg/kg body weight basis, extrapolated from the phase I study of DMA as an anti-cancer agent [10]. The description of the IV-Bu serious advers event (SAE) profile is strikingly similar to that experienced by all ten patients who received DMA at a total dose of 400 mg/kg/day for at least 3 days in its phase I study [14]. Three of the ten patients at the highest dose level developed hypotension and a shock-like syndrome that cleared in two of them, although the third patient subsequently developed "meningismus" with a negative work-up for infection and expired [14]. Because of the known association of DMA with serious toxicities in animals and humans it is considered a class II agent, i.e., its use in pharmaceutical formulations should be very limited, or preferably avoided [13,25,26]. In fact, the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use, Step 4, currently limits the daily exposure of DMA to 10.9 mg [26]. This recommended maximal exposure is exceeded by at least 1000-1500-fold when adults receive a myeloablative 4-day course of IV Bu/DMA with Cy or Flu [13,[22][23][24]. We herein report six patients who experienced hallucinations, agitation, delirium, and deep stupor in one patient lasting for almost 2 months before clearing. One additional patient developed stupor, which advanced into encephalopathy, irreversible coma and death. Those cases illustrate that it may be advisable to heed the FDA expert reviewer's concern and exercise caution with the DMA-containing Bu formulation, Busulfex™, at least in patients with preexisting CNS history of radiotherapy or cerebrovascular disease.
Further, OMI claimed that the maximum DMA dose delivered with their BuCy2 regimen was about 42% of that reported toxic by Weiss [14]; however, a more careful calculation shows that the dose given over 4 days exceeds 60%. In addition, many centers have moved from dosing based on body size and utilize pharmacokinetic (PK) dose guidance for IV Bu [21,23,24]. Due to interindividual heterogeneity in human drug metabolism, patients who rapidly metabolize busulfan may now receive IV Bu doses ≥200 mg/m 2 . This translates into a DMA dose >80% of the dose reported by Weiss as toxic [14]. Therefore, we suggest that there may not only be a risk for DMA-associated toxicity, but also a potential for serious clinical interactions between DMA and Bu, and with other components of multiagent regimens, which may be further obfuscated by interindividual differences in drug metabolism.

CONCLUSION
We suggest that caution should be exercised with use of IV Bu in patients who have an underlying organic brain disorder, or who have received brain XRT or immunoconjugates with a potential for causing CNS adverse events.

DATA AVAILABILITY
Raw data were generated at the University of Texas MD Anderson Cancer Center. Derived data supporting the findings of this study are available from the corresponding author JLR upon reasonable request.