Electroencephalogram Abnormality and High-Dose Busulfan in Conditioning Regimens for Stem Cell Transplantation

Electroencephalogram abnormality and high-dose busulfan in conditioning regimens for stem cell transplantation

R Kobayashi, N Watanabe, A Iguchi, Y Cho, M Yoshida, H Arioka, H Naito, T Shikano and Y Ishikawa

Department of Pediatrics, Hokkaido University School of Medicine, Sapporo, Japan

Summary: Patients and methods

High-dose busulfan (BU) is widely used in combined Between June 1989 and November 1996, a total of 22 chemotherapy before allogeneic or autologous bone patients with various hematological malignancies or marrow transplantation. Convulsions are reported as a immunodeficiencies received allogeneic BMT (14 cases) or side-effect of high-dose BU. We recorded electroence- autologous peripheral blood stem cell transplants (eight phalograms (EEGs) before and on the third day of BU cases) after preparation with high-dose BU (Table 1). Thir- administration in 22 patients. Abnormal EEGs were teen patients had acute lymphoblastic leukemia (ALL): observed on the third day in 13 cases (59%). These eight were in first complete remission (CR) and five were patients were older (P Ͻ 0.05) and had had larger doses in second CR. Five patients had acute non-lymphoblastic of BU (P Ͻ 0.025) than the nine patients with normal leukemia (ANLL): three were in first CR and two were EEGs. Convulsions occurred in two of the 22 patients, in second CR. Five patients had other diseases: two non- one of whom was receiving prophylaxis with phenytoin. Hodgkin’s lymphoma (NHL), one Wiskott Aldrich syn- Gamma aminobutyric acid (GABA), a natural mediator drome (WAS), and one chronic myelogenous leukemia of defense against epileptic activity, concentrations in (CML). cerebrospinal fluid measured before and after adminis- The conditioning regimen used consisted of BU given tration of BU showed no definite changes. orally over 4 consecutive days, followed by cyclophos- Keywords: busulfan; convulsion; electroencephalogram; phamide (CY) and/or other drugs. The dose of BU was GABA 16 mg/kg in older children and 500 mg/m2 in younger children. Every patient except when otherwise stated received anticonvulsant prophylaxis with phenytoin (4 mg/kg intravenously as a loading dose followed by 4 mg/kg/day High-dose busulfan (BU) is increasingly used in condition- orally). Phenytoin administration was begun 12 h before the ing regimens for bone marrow transplantation (BMT). first dose of BU and continued until at least 24 h after the Liver dysfunction is reported as a side-effect of high-dose last dose. BU. Since the introduction of high-dose BU in conditioning EEGs were recorded before, and on the third day of BU regimens for bone marrow transplantation, several cases of administration and were evaluated by pediatric neurol- 1–6 convulsions have been reported, especially in adults. In ogists. The administration of BU and phenytoin, and sched- the literature, convulsion occurs in adults after the second ule for EEG recording are illustrated in Figure 1. GABA day of therapy, more frequently after the last dose of BU. concentrations in cerebrospinal fluid samples were meas- Phenytoin is usually given to adult patients receiving high- ured in 12 of 22 patients, four with normal EEGs and eight dose BU, but convulsions still occur after prophylaxis. In with abnormal EEGs before and on the third day of BU children, the acute neurotoxicity of BU is reported to be administration using high performance liquid chromatogra- 4 dose dependent. Vassal et al reported a significant differ- phy (HPLC). ence between children receiving 16 mg/kg and those receiv- A t test or Wilcoxon test was used to compare age and 2 ing 600 mg/m in terms of probability of developing acute doses of BU. The paired t test was used to compare the neurotoxicity. However, several studies are available which concentration of GABA before and after administration examined the EEG as a possible index of neurotoxicity dur- of BU. ing BU administration.7 We recorded electroencephalo- grams and measured gamma aminobutyric acid (GABA), a natural mediator of defense against epileptic activity, from Results cerebrospinal fluid samples taken before and on the third day of BU. EEGs before administration of BU were normal in 20 of 22 patients. One of the two patients with an abnormal EEG before BU had leukoencephalopathy diagnosed by MRI, Correspondence: R Kobayashi, Department of Pediatrics, Hokkaido Uni- associated with cranial radiation and methotrexate. EEGs versity School of Medicine, North 15-West 7, Kitaku Sapporo, 060, Japan performed on the third day of BU administration were Received 3 June 1997; accepted 25 August 1997 abnormal in 13 of 22 patients. Abnormal EEGs showed EEG abnormality and high-dose busulfan R Kobayashi et al 218 Table 1 Patient characteristics

Case Age Sex Type of SCT Disease Term to Treatment Past history Seizure EEG Dose of busulfan (years) SCT of cranial abnormality irradiation (mg/kg) (mg/m2)

1 15 F AlloBMT ALL CR2 15 BU+VP16+AraC+CY −++16 540.5 2 3 F AlloBMT CML CP 28 BU+CY −−−16 382.1 3 9 M AlloBMT ALL CR2 67 BU+VP16+CY +−−16 470.6 4 16 M AlloBMT ANLL CR1 10 BU+CY −++16 509.6 5 10 M AutoPBSCT ALL CR2 60 BU+VP16+CY +−−16 508.5 6 1 M AlloBMT ALL CR1 8 BU+VP16+CY −−−18 425.5 7 12 F AlloBMT NHL CR1 8 BU+VP16+CY −−−16 472.4 8 1 F AutoPBSCT ALL CR1 12 BU+VP16+CY −−−20 500.0 9 2 F AlloBMT ALL CR1 9 BU+VP16+CY −−−20 500.0 10 16 M AutoPBSCT ALL CR1 11 BU+VP16+CY −−+16 557.6 11 14 M AlloBMT ANLL CR2 72 BU+L-PAM +−+16 538.5 12 15 M AutoPBSCT ALL CR1 8 BU+VP16+CY −−+16 500.0 13 3 F AutoPBSCT ALL CR1 8 BU+VP16+CY −−+20 483.9 14 4 M AlloBMT WAS 51 BU+ATG+CY −−+22.6 491.2 15 15 F AutoPBSCT ALL CR1 6 BU+VP16+CY −−+16 486.5 16 13 M AlloBMT ANLL CR2 27 BU+L-PAM+ATG +−−20 500.0 17 6 F AlloBMT ALL CR2 58 BU+VP16+CY −−+16.8 487.8 18 14 M AutoPBSCT NHL CR1 10 BU+VP16+CY −−+16 582.3 19 10 F AutoPBSCT ALL CR2 56 BU+VP16+CY +−−16 516.1 20 4 M AlloBMT ALL CR1 11 BU+VP16+CY −−+20 500.0 21 6 M AlloBMT ANLL CR1 5 BU+L-PAM −−+21.4 500.0 22 9 M AlloBMT ANLL CR1 7 BU+L-PAM −−+20.6 500.0

SCT = stem cell transplantation; BMT = bone marrow transplantation; PBSCT = peripheral blood stem cell transplantation; ALL = acute lymphoblastic leukemia; ANLL = acute non-lymphoblastic leukemia; CML = chronic myelogenous leukemia; NHL = non-Hodgkin’s lymphoma; WAS = Wiskott Ald- rich syndrome; BU = busulfan; CY = cyclophosphamide; VP16 = etoposide; AraC = cytosine arabinoside; L-PAM = melphalan; ATG = antithymocyte globulin; CP = chronic phase; CR = complete remission.

days 12345 myelogenous leukemia in first complete remission who Busulfan received BU while taking phenytoin. On the third day of 1~1.25mg/kg po BU administration his EEG revealed polyspikes without Phenytoin convulsions. Twenty hours after the last dose of BU, he 4 mg/kg i.v. suffered a generalized tonic–clonic seizure of 3 min dur- Phenytoin ation with loss of consciousness, and received intravenous 2 mg/kg po diazepam. He had no other complication, including liver dysfunction, and is alive and disease-free. EEG Compared to patients with normal EEGs, patients with Figure 1 Administration of busulfan and phenytoin, and schedule for abnormal EEGs were older in age (P Ͻ 0.05) and received electroencephalogram. Phenytoin, 4 mg/kg intravenously as a loading dose larger doses of BU (P Ͻ 0.025) (Table 2). Sex, disease and followed by 4 mg/kg/day oral administration, was begun 12 h before the past history of cranial irradiation were not associated with first dose of busulfan and continued until at latest 24 h after the last dose. EEGs were recorded before and on the third day of busulfan. EEG abnormalities. GABA concentrations measured in cerebrospinal fluid decreased on the third day of BU administration in five of diffuse polyspikes in one patient and diffuse spike-and- 12 patients tested including those with and without EEG waves in 12 patients. An example is shown in Figure 2. abnormalities. In the remaining seven patients, GABA The EEGs of the two patients with abnormalities before increased, and thus, no definite tendency was observed in BU worsened after administration of BU. the changes of GABA in CSF between patients with and Convulsions occurred in two patients. One was a 15- without EEG abnormalities (Figure 3). year-old girl with acute lymphoblastic leukemia in second remission who received BU without administration of anti- Discussion convulsant prophylaxis. On the third day of BU administration (after the 8th dose), she suffered a generalized The mechanism of neurotoxicity with BU is not clear. tonic–clonic convulsion of 2 min duration with loss of con- However, it is known that BU does not bind tightly to sciousness which resolved spontaneously. No abnormality plasma proteins and easily crosses the blood–brain barrier was noted on brain CT. Her EEG after the convulsion soon after oral administration, and that CSF concentration revealed spontaneous spike-and-wave complexes. She was is similar to, and sometimes significantly higher than that given oral phenobarbital treatment thereafter but died with in plasma. Thus, convulsions occurring after the last dose hepatic veno-occlusive disease (VOD) on day +56 post- of BU could be considered to be due to prolonged brain BMT. The other case was a 16-year-old boy with acute exposure to BU and/or to its metabolites. EEG abnormality and high-dose busulfan R Kobayashi et al 219 Before busulfan Third day of busulfan

Fp1 Fp1

Fp2 Fp2

C3 C3

C4 C4

P3 P3

P4 P4

O1 O1

O2 O2

F7 F7

F8 F8

T5 T5 µ 50 V µ 1 sec 50 V T6 T6 1 sec

Figure 2 An example of EEG taken for case 20 before and on the third day of busulfan administration. Spike and wave complexes were seen on the third day of busulfan.

Table 2 Proﬁle of cases with or without EEG abnormality

Abnormal EEG (n = 13) Normal EEG (n = 9)

Sex (M/F) 9/4 4/5 Age (years, mean ± s.e.) 10.54 ± 1.45* 6.78 ± 1.65* Disease ALL 7 ALL 6 ANLL 4 ANLL 1 NHL 1 NHL 1 WAS 1 CML 1 Past history of cranial irradiation 1 1 Dose of busulfan (mg/kg, mean ± s.e.) 17.95 ± 0.70 17.56 ± 0.65 (mg/m2, mean ± s.e.) 513.68 ± 8.60** 475.02 ± 14.87**

*P Ͻ 0.05; **P Ͻ 0.025.

We noted two cases of BU-associated convulsions in 22 barbital, clonazepam and phenytoin are usually given to BMT recipients conditioned with high-dose BU. The over- patients receiving high-dose BU. However, even with the all incidence of convulsions was 9.1%. However, in 13 out use of phenytoin, some patients are reported to suffer con- of 22 cases (59%), abnormal EEGs were recorded after vulsions. It is known that serum concentrations of pheny- administration of BU. Since almost all patients received toin increase gradually with oral administration, and may prophylaxis with phenytoin, this prophylactic procedure sometimes not assume high enough levels to prevent may be effective in preventing convulsions due to BU. We convulsions due to BU. In this context, we recommend that compared factors between patients with and without abnor- the first dose of phenytoin should be administered mal EEGs. EEG abnormalities correlated with age, and intravenously. dose of BU confirming an earlier report which showed that Gamma aminobutylic acid (GABA), a major neuro- neurotoxicity with BU is dose-dependent in children.4 transmitter in the brain, may be a natural mediator for Recently it was shown that 16 mg/kg of BU was insuf- defense against epileptic activity. It has been speculated ficient for conditioning in younger children and larger doses that decreases in GABA are associated with convulsions of BU were recommended. However, as is shown in the due to penicillins, cephalosporins, and procaine.8 Further- present study, a high dose of BU promotes EEG abnor- more, concentrations of GABA in cerebrospinal fluid have mality, a possible indicator of neurotoxicity. Thus, con- been reported to be low in patients suffering from convulsion prophylaxis is important for these patients. Pheno- vulsions.9 We measured GABA in cerebrospinal fluid EEG abnormality and high-dose busulfan R Kobayashi et al 220 1200 sidered when administering larger doses of BU to children, especially older ones. 1000 Acknowledgements 800 We thank Prof Kunihiko Kobayashi for helpful discussion.

600 References 400 GABA (pmol/ml) 1 Marcus R, Goldman JM. Convulsion due to high dose busulfan. Lancet 1984; ii: 1463. 200 2 Martell RW, Sher C, Jacobs P, Monteagudo F. High-dose busulfan and myoclonic epilepsy (letter). Ann Intern Med 1987; 0 106: 173. 3 Camara RDL, Tomas JF, Barberana M, Fernandez-Ranada JM. before after High-dose busulfan and seizures. Bone Marrow Transplant 1991; 7: 363–364. administration of busulfan 4 Vassal G, Deroussent A, Hartmann O et al. Dose-dependent abnormal EEG neurotoxicity of high-dose busulfan in children: a clinical and pharmacological study. Cancer Res 1990; 50: 6203–6207. normal EEG 5 Grigg AP, Shepherd JD, Philips GL. Busulfan and phenytoin. Figure 3 Change in GABA concentration in cerebrospinal fluid before Ann Intern Med 1989; 111: 1049–1050. and after busulfan. GABA concentration measured in cerebrospinal fluid 6 Sureda A, Perez de Oteyza J, Garcia Larana J, Odriozola J. decreased on the third day of BU administration of five of 12 patients High dose busulfan and seizures. Ann Intern Med 1989; 111: tested including those with and without EEG abnormality. In the remain- 543–544. ing seven patients, GABA increased, and thus no definite tendency was 7 Meloni G, Raucci U, Pinto M et al. Pretransplant conditioning observed in the changes of GABA in CSF between patients with and without EEG abnormality. with busulfan and cyclophosphamide in acute leukemia patients: neurological and electroencephalographic prospective study. Ann Oncol 1992; 3: 145–148. samples before and on the third day after BU, and found 8 Sawaki K, Ouchi K, Sato T, Kawaguchi M. Some correlations between local anesthetic-induced convulsions and gamma-ami- no definite trends in GABA levels on the two occasions, nobutylic acid in rat spinal fluid. Jpn J Pharmacol 1991; 56: possibly indicating that GABA concentration is not respon- 327–335. sible for convulsion after BU. 9 Wood JH, Hare TA, Glaeser BS et al. Low cerebrospinal fluid The mechanism of convulsion after BU is unclear. ␥-aminobutylic acid content in seizure patients. Neurology Nevertheless, the possibility of neurotoxicity must be con- 1979; 29: 1203–1208.