Original Contribution

Incidence and onset of delayed seizures after overdoses of extended-release bupropion

Paul Starr PharmD^a,?, Wendy Klein-Schwartz PharmD^a, Henry Spiller MS, RN^b,

Perri Kern BSN, RN^c, Susan E. Ekleberry BS Pharmacy, RPh^d, Susan Kunkel PharmD^e

^aMaryland Poison Center/Department of Pharmacy Practice and Science,

University of Maryland Baltimore, Baltimore, MD 21201

^bKentucky Regional Poison Center, Kosair Children’s Hospital, Louisville, KY 40202 ^cRegional Poison Control Center, Children’s Hospital of Michigan, Detroit, MI 48201 ^dCentral Ohio Poison Center, Nationwide Children’s Hospital, Columbus, OH 43205

^eNew Mexico Poison and Drug Information Center, University of New Mexico Health Sciences Center, Albuquerque, NM 87131

Received 7 April 2008; revised 29 June 2008; accepted 2 July 2008

Abstract

Background: Delayed seizures have been reported with overdoses of bupropion extended-release (XL). This study systematically evaluates the frequency and timing of seizures and an association between other toxic effects (ie, agitation, tremors, and hallucinations) and seizures.

Methods: A 3-year multi-poison center observational study of hospitalized patients with ingestion of bupropion XL >=600 mg in adults and >=4 mg/kg in children was performed. Patients with coingestants or a medical history that could affect seizure occurrence were excluded. Data collection forms captured onset time of seizure(s), other symptoms, and treatment.

Results: One hundred seventeen patients met inclusion criteria: median age of 22 years (range, 1.3-65 years) with 16 children <= 3 years. Seizures occurred in 37 (31.6%) patients, with initial seizure at 0.5 to 24 hours after ingestion; 12 (32%) patients had initial seizure at N 8 hours. Subsequent seizures occurred in 49%. Children ages 1.3, 3, and 7 years, developed seizures. In patients >= 13 years of age, median dose with seizures was 4350 mg (range, 600-54 000) compared to 2400 mg (range, 600-9000) in patients without seizures. Agitation, tremors, and hallucinations occurred in 29.7%, 40.5%, and 18.9% of patients with seizures, respectively, compared with 12.5 %, 17.5%, and 10% in patients without seizures. The neurologic effects agitation (P = .045) and tremors (P = .005) occurred more frequently. Conclusion: Delayed seizure onset suggests a minimum observation period of 24 hours after bupropion XL overdose. Although patients experiencing agitation or tremors may be at greater risk, seizures can occur without preceding Central nervous system toxicity.

(C) 2009

* Corresponding author. Maryland Poison Center, University of Mary- land School of Pharmacy, Level 01, Baltimore, Maryland 21201.

E-mail addresses: [email protected], [email protected] (P. Starr).

Introduction

Bupropion is indicated for depression and as adjunctive therapy in smoking cessation. Bupropion is less likely to cause weight gain and sexual dysfunction compared to other

0735-6757/$ - see front matter (C) 2009 doi:10.1016/j.ajem.2008.07.004

commonly prescribed antidepressants. It was removed from the market shortly after introduction in the mid 1980s because of seizures as an adverse effect but was reintroduced once it was determined that the frequency of seizures drastically reduced with daily doses of 450 mg or less [1]. Currently, it is available in immediate release (IR), sustained release (SR), and extended release (XL) formulations marketed under the proprietary names of Wellbutrin and Zyban (GlaxoSmithKline, Research Triangle Park, NC). The XL formulation, approved in 2003 and marketed in 2004, is available in 150- and 300-mg tablets taken once daily. The dose can be titrated up to a maximum dose of 450 mg.

Bupropion is chemically characterized as a monocyclic aminoketone unlike any other antidepressant on the market. The active metabolites R,R-hydroxybupropion, S,S-hydro- xybupropion, threo-hydrobupropion, and erythro-hydrobu- propion result from extensive liver metabolism. blood levels of bupropion XL peak approximately 5 hours after ingestion. Both the parent compound and active metabolites follow linear kinetics [2]. Area under the curve concentrations are higher, and elimination half-lives of the active metabolites are longer compared to bupropion. There is documented interindividual variability [1,3]. The mechanism of action is thought to be a weak inhibition of the reuptake of dopamine and norepinephrine. It is not known to interfere with serotonin reuptake or inhibit monoamine oxidase [2,4].

A narrow therapeutic margin is evident from observa- tional studies that show Seizure activity with doses of 600 mg or higher. Seizure activity occurred in 21% of patients with overdoses of the IR formulation and 11% of patients described in a study composed predominately of overdoses of the SR formulation [5,6]. Delayed onset of seizure activity occurred up to 8 hours (mean, 3.7 hours) and 14 hours (mean, 4.3 hours) with overdoses on IR and SR, respectively [5,6]. cardiac symptoms other than tachycardia rarely occur with bupropion overdose. Hallucinations and agitation are associated with overdoses of both the IR and SR formula- tions [7]. Tremors occurred in 7.1% to 24% of observed patients exposed to bupropion supraTherapeutic doses [8-10]. There are no studies that determine the onset of seizure activity in patients with overdoses of the XL formulation. Therefore, information is lacking upon which to base observation time guidelines for these patients when they are managed in an emergency department (ED). The objectives of this observational study are to determine a minimal observation period and whether the occurrence of other toxic

manifestations can help predict seizure occurrence.

Methods

A 1-year retrospective chart review (January 1, 2005- December 31, 2005) and 2-year prospective observational study (January 1, 2006-December 31, 2007) of cases from 5 poison centers was performed. Participating centers were the Children’s Hospital of Michigan Poison Center, Central

Ohio Poison Center, Kentucky Regional Poison Center, Maryland Poison Center, and the New Mexico Poison and Drug Information Center. All cases were documented in the respective poison centers’ online data entry systems. All the poison center online data entry systems uniformly include the reported time and amount of ingestion, symptoms reported, and outcome. In addition, a standardized data collection form that captured the occurrence and timing of toxic manifesta- tions and the treatment of each specific symptom was completed. A coinvestigator at each center was responsible for reviewing the cases to ensure they met inclusion criteria before submission. Final eligibility for inclusion was determined by the primary investigator after rigorous auditing of each case. Patients of all ages were included with Acute ingestions 600 mg or greater in patients 12 years or older and 4 mg/kg or greater in patients younger than 12 years. If the exact dose was unknown but the minimum estimated dose ingested met the dose inclusion criterion, the case was included. If the weight was not known in patients younger than 12 years, the milligram per kilogram dosage was determined by imputing the weight using previously described methods [11]. All patients were required to have been observed in a hospital setting for a minimum of 24 hours after ingestion with clinical outcome documented. If the exact time of ingestion was unknown or estimated, observation for a period no less than 24 hours from the latest estimated time of ingestion was required for inclusion.

Exclusion criteria included a history of coingesting an unknown substance or any substance known to lower seizure threshold or suppress seizures. Patients with a prior history of seizures including those currently being treated with antic- onvulsants for seizures or as a mood stabilizer were excluded. Reported physical dependence on any substance that could result in seizure if abruptly stopped (eg, alcohol) resulted in excluding the case. Routine urine drug screens positive for cocaine or benzodiazepines (before treatment) resulted in exclusion. A positive urine drug screen for amphetamines was also excluded unless it was confirmed that no medications or amphetamines were ingested pointing to a “false-positive” caused by the bupropion [2]. Any case not managed in a health care facility was excluded. Patients who left against medical advice before the 24 hour postingestion period as well as any case without a known outcome were automatically excluded.

The data evaluated included demographics such as age and sex, dosage, coingestants, presence or absence of other symptoms associated with bupropion toxicity, onset time of seizures, treatment, and outcomes. medical outcomes were defined in accordance with the American Association of Poison control centers National Poison Data System as: no effect (no signs or symptoms as a result of the exposure), minor effect (signs or symptoms that were minimally bothersome and resolved rapidly), moderate effect (signs or symptoms that were more pronounced, more prolonged, or more systemic in nature than minor, usually requiring treatment but not life-threatening), major effect (signs or

symptoms that were life-threatening or resulted in significant residual disability), or death (death resulted from the exposure or a direct complication of the exposure) [12]. Patients with seizures were compared to nonseizure patients with regard to dose and the presence of other signs of toxicity, specifically agitation, tremors, hallucinations, and tachycardia (heart rate N100 beats per minute). Patients who had seizures were evaluated for time of onset of their first and subsequent seizures. Data were extracted from the poison center record and the data collection form and input into Excel for analysis. Data were analyzed using VassarStats (Lowry R, http://faculty.vassar.edu/lowry/VassarStats.html; 1998-2008). Statistical tests included the Mann-Whitney nonparametric test to compare dosages ingested by the seizure patients and noneizure patients. ?2 and Fisher exact tests were performed to analyze for differences between the 2 groups with regard to frequency of other toxic manifestations of bupropion overdose. Sensitivity, specificity, and positive and negative predictive values were determined to evaluate the extent to which common clinical effects could predict the presence or absence of seizures.

The study was approved by the institutional review boards at each of the participating poison centers.

Results

After initial screening at the originating centers, 427 cases were submitted for review to the investigators at the coordinating poison center. Of these cases, 117 patients met the criteria for inclusion and comprise the study group. The study group includes 53 cases identified retrospectively and 64 identified prospectively. There were 71 females (60.6%) and 46 males (39.4%). Ages ranged from 16 months (1.3 years) to 65 years (mean, 25.8 +- 15.8; median, 22). There were 16 children 3 years or younger, one 7-year-old, 32 adolescents (13-19 years of age), and 68 adults.

Bupropion was the only reported substance ingested in 97 (82.9%) patients. Ethanol was a coingestant in 7 cases but had no effect on occurrence of seizures (P = .21). Mean and median doses by history in adolescents and adults were 4317.6 +- 6525.7 and 3000 mg, respectively (12 unknown dose). Mean and median dosages in children 7 years or younger ranged were 35.6 +- 42.3 and 19.3 mg/kg, respectively (2 unknown dosage).

Seizures occurred in 37 (31.6%) patients (32% retro- spectively, 31% prospectively) with onset time of the initial seizure ranging from 0.5 to 24 hours after the overdose (mean, 7.3 +- 5.4; median, 6.1; retrospective mean, 7.9 +-

1. 8; prospective mean, 7.0 +- 5.3). The patient who experienced a first seizure at 24 hours after ingestion was in the prospective group. The timing of the initial seizure was unknown in 3 cases. In 25 patients, the initial seizure occurred in 8 hours or less. Initial seizures were delayed 9 or more hours in 7 (24.3%) patients. In these patients, time of first seizure after the overdose was 11 to 11.25 hours in 2

patients, 12 hours in 2 cases, 14 hours in 1 case, 15 hours in

3 cases, and 24 hours in 1 case. Of the 37 patients who seized, more than one seizure was reported in 18 (49%) patients, with the average onset time for the second seizure of 9.8 hours postingestion (median, 9.5) with a range of 1.5 to 19 hours. The second seizure was reported an average of

2.6 hours after the first seizure (range, immediate onset to 14 hours). For 7 patients, the second seizure occurred within

1 hour of the first seizure. The timing of subsequent seizures in the 8 patients experiencing 3 or more seizures was 4 to 22 hours after ingestion with a mean of 13.8 hours. Timing of the first seizure did not influence the frequency of second seizures (P = .69).

Adolescents and adults with seizures ingested larger doses compared to patients without seizures (Table 1). There was no relationship between dose and timing of initial seizure (P = .65). Only 3 of 17 children 7 years or younger developed seizures. Dosages between children with and without seizures were not significantly different (Table 1).

Evaluation of other clinical effects associated with bupropion overdose found that tachycardia, agitation, and tremors were more frequent in patients with seizures (Table 2). The frequency of hallucinations was not significantly different for seizure and nonseizure patients. In comparing the number of common neurologic symp- toms (agitation, tremor, and hallucination), seizure patients were more likely to experience any 1 or 2 clinical effects than nonseizure patients (P = .037 for both 1 and 2 symptoms). There was no difference in seizure frequency in patients experiencing all 3 neurologic effects (P = .088). In all, 17 seizure patients (46%) did not exhibit agitation, tremor, or hallucinations (Table 3). The sensitivity, specificity, positive predictive values and negative pre- dictive values for tachycardia and the neurologic effects are presented in Table 4.

Table 1 Age and dose comparison of seizure vs nonseizure patients by age group

Adolescents and adults	Seizure patients (n = 34)	Nonseizure Patients (n = 66)	P
Age (y)
Median (range)	21.5	29
Mean +- SD	26.8 +- 12.3	31.2 +- 14.0	.16
Dose (mg)
Median (range)	4350 (600-54 000)	2400 (600-9000)
Mean +- SD	7145 +- 10 383	2800 +- 1930	.015
Patients <=	Seizure	Nonseizure	P
7 y of age	patients (n = 3)	Patients (n = 14)
Age (y)
Median (range)	3	2
Mean +- SD	3.8 +- 2.9	2.4 +- 0.5	.61
Dosage (mg/kg)
Median (range)	18.7 (16-143)	19.4 (10.7-133)
Mean +- SD	59.2 +- 72.5	29.7 +- 33.6	.29

Clinical effect	Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)
Tachycardia	91.2	48.8	45.3	92.9
Agitation	29.7	87.5	52.4	72.9
Tremors	21.6	85.0	55.6	75.6
Hallucinations	18.9	90.0	46.7	70.6
PPV indicates positive predictive value; NPV, negative predictive value.

Other documented clinical effects are listed in Table 5. In 27 cases, patients remained asymptomatic throughout the observation period of at least 24 hours after ingestion. Coded medical outcomes in the remaining patients were 32 minor effects, 42 moderate effects, 15 major effects, and 1 death. The death occurred in a 16-month-old boy who ingested 143 mg/kg of bupropion. Ventricular fibrillation, documented after several periods of repetitive seizure activity, preceded cardiac arrest and death 6.5 hours after the overdose.

Table 2 Comparison of associated clinical findings and treatment, seizure vs nonseizure patients

Gastrointestinal decontamination, consisting of activated charcoal (n = 81) and/or whole bowel irrigation (n = 30) and/ or lavage (n = 6), was performed in 85 (72.6%) patients. Gastrointestinal decontamination did not impact seizure frequency (P = .132) (Table 2). Benzodiazepines and/or other anticonvulsants were administered to 44 patients, 30 seizure patients and 14 nonseizure patients (Table 2).

Discussion

In this cohort of bupropion XL overdoses, seizures developed in 32% of patients, a considerably higher frequency than the previously reported 21% with the IR and 11% with the SR formulations of bupropion [5,6]. Our findings are strengthened by the fact that we excluded

patients with coingestants and/or medical conditions that alter seizure threshold. The higher frequency may in part be explained by the minimum dose inclusion criteria of 600 mg in adults and 4 mg/kg in children. In Spiller’s study of 58 patients with overdoses of the IR product without coin- gestants, doses ranged from as low as 200 up to 6300 mg, with the lowest dose resulting in seizures being 575 mg [5]. In Shepherd’s study, in which most of the 385 patients ingested the SR formulation, there is no information on the coingestants that were involved in 37% of the overdose patients that seized. Doses in seizure patients were 600 to 18 g; doses in nonseizure patients were not reported.

Seizure onset ranged from 0.5 hours up to 24 hours after ingestion (mean, 7.3 hours) with one fourth of the initial seizures occurring more than 8 hours after ingestion. This compares to previous studies in which patients experienced seizures from the IR and SR formulations as late as 8 and 14 hours, respectively. For both the IR and SR formulations, the mean time to seizure onset was close to 4 hours after the overdose. Patients overdosing on the XL bupropion product demonstrated an 80% increase in the average time of onset of seizures and a markedly delayed time frame for late presenting seizures compared with the IR and SR formula- tions. Approximately half the patients who had a seizure after bupropion XL overdose experienced more than one seizure. Based on these findings, an observation period of at least 24 hours is recommended.

Associated clinical findings	Seizure patients (n = 37)	Nonseizure patients (n = 80)		P
Tachycardia	34 (91.8%)	41	(51.2%)	.00005
Agitation	11 (29.7%)	10	(12.5%)	.045
Hallucinations	7 (18.9%)	8	(10.0%)	.29
Tremor	15 (40.5%)	12	(17.5%)	.005
Treatment
No. with GI	23 (62.1%)	62	(77.5%)	.13
decontamination
No. with	30 (81.0%)	14	(17.5%)	b.0001
benzodiazepines
and/or anticonvulsants

Defining the toxic dose is difficult. In this study, adolescent and adult patients who seized ingested higher doses than those without seizures. However, seizures

Table 5 Other related clinical effects

Table 4 Correlation of common clinical effects with seizures

Table 3 Relationship between number of CNS effects and seizures ^a

Seizures (n = 37)

No seizures (n = 80)

P a

1 CNS effect

27%

2 CNS effects 3 CNS effects

19% 8.0%

14%

8.0%

2.5%

.037 ^b

.037 ^b

0.088 ^c

^a P values based on number of symptoms compared to no symptoms.

^b ?² Test.

^c Fisher exact test.

Toxic effects	No. of cases (all patients) (n = 117)	No. of seizure patients (n = 37)
Nausea and/or vomiting	21 (18%)	4 (11%)
Hypertension	12 (10%)	3 (8%)
Drowsiness	10 (9%)	5 (12%)
Dizziness	6 (5%)	2 (5%)
Hypotension	5 (4%)	1 (3%)
Coma	2 (2%)	2 (5%)
Confusion	2 (2%)	1 (3%)
Acidosis	2 (2%)	2 (5%)
Conduction	2 (2%)	2 (5%)
delays/dysrhythmias

occurred at the minimum inclusion dose of 600 mg. Given the wide range of doses in both seizure and nonseizure patients, dose alone should not be the deciding factor in determining the clinical observation period.

Agitation and tremor occurred more often in the seizure group. Tachycardia was also reported significantly more frequently in seizure compared to nonseizure patients. Presence of 1 or 2 of the common neurologic effects (agitation, tremor, and hallucination) of bupropion toxicity may help predict a patient’s propensity toward seizures. Tachycardia has the highest sensitivity and lowest specificity as well as the highest negative predictive value. Therefore, tachycardia is most likely to be present in patients who seize, whereas in the absence of tachycardia, seizures are unlikely. Specificity and negative predictive values are high for all 3 neurologic effects such that if the neurologic effect is absent, there is a low likelihood of seizures. However, 17 patients experienced seizures without exhibiting any of these neurologic effects, suggesting that the patient’s clinical neurologic presentation may not be an accurate predictor of subsequent seizure activity.

Gastrointestinal decontamination was frequently per- formed and consisted primarily of activated charcoal and/ or whole bowel irrigation. Given the extended release nature of this bupropion product, GI decontamination can poten- tially decrease gastrointestinal absorption, even when used hours after the overdose. This study was unable to demonstrate any impact of GI decontamination on seizure frequency. Some patients experienced a seizure before GI decontamination, making it difficult to assess whether GI decontamination influenced outcome. Although it is con- ceivable that these patients might have experienced sub- sequent seizures if GI decontamination had not been performed, it is not possible to evaluate this scenario with these data.

This study has several limitations. The nature of observational studies based on documented poison center cases opens a study to a number of opportunities for possible inaccuracy [13]. Retrospective and prospective cases were included to increase the number of subjects in our study population that could compromise accuracy of data. However, comparison of retrospective cases with prospective cases showed similar frequency and onset of seizures. In some cases, patients presented to the ED after the initial seizure occurred, leaving some question as to the exact time of onset of seizure activity. However, this study used a standardized data collection form for all cases with a rigorous review of each case to minimize any inconsis- tencies or omissions that may affect the accuracy of data. Ingested amounts, time of ingestion, and onset of symptoms were based on reports by parents, patients, and health care providers. Because bupropion blood levels are

not routinely monitored, there was no laboratory confirma- tion of the overdose.

Conclusions

All patients with a suspected overdose of bupropion XL should be monitored for in a health care facility for a minimum of 24 hours after the overdose. In adolescents and adult patients, doses 600 mg or higher of the extended release formulation can cause seizures. In children, the lowest dose associated with seizures was 16 mg/kg. Given the small number of children with seizures in our case series, the minimum toxic dose is less well defined in this population. Patients presenting with agitation, tremor, or tachycardia may be at greater risk for seizure activity but seizure activity can occur without prior toxic neurologic effects.

References

1. Davidson J. Seizures and bupropion: a review. J Clin Psychiatry 1989;50(7):256-61.
2. Product Information: WELLBUTRIN XL(R) extended release oral tablets, bupropion hcl extended release oral tablets. Research Triangle Park, NC: GlaxoSmithKline; 2006.
3. Hesse LM, He P, Krishnaswamy S, Hao Q, Hogan K, von Moltke LL, Greenblatt DJ, Court MH. Pharmacogenetic determinants of inter- individual variability in bupropion hydroxylation by Cytochrome P450 2B6 in human liver microsomes. Pharmacogenetics 2004;14 (4):225-38.
4. Stahl S, Pradko J, Haight B, Modell J, Rockett C, Learned-Coughlin S. A review of the neuropharmacology of bupropion, a dual norepinephr- ine and dopamine reuptake inhibitor. Prim Care Companion J Clin Psychiatry 2004;6(4):159-66.
5. Spiller HA, Ramoska EA, Krenzelok EP. Bupropion overdose: a 3-year multi-center retrospective analysis. Am J Emer Med 1994;12:43-5.
6. Shepherd G, Velez LI, Keyes DC. Intentional bupropion overdoses. J Emerg Med 2004;27(2):147-51.
7. POISINDEX System [database on CD-ROM]. Version 5.1. Green- wood Village, Colo: Thomson Healthcare.
8. Colbridge MG, Dargan PI, Jones AL. Bupropion-the experience of the National Poisons Information Service (London). J Toxicol Clin Toxicol 2002;40(3):398-9.
9. Belson MG, Kelley TR. Bupropion exposures: clinical manifestations and medical outcome. J Emerg Med 2002;23(3):223-30.
10. Shepherd G. Adverse effects associated with extra doses of bupropion.

Pharmacotherapy 2005;25(10):1378-82.

1. Benson, Spyker DA, Troutman WG, Watson WA. TESS-based dose- response using pediatric clonidine exposures. Toxicol Appl Pharmacol 2006;213:145-51.
2. Bronstein AC, Spyker DA, Cantilena Jr LR, Green J, Rumack BH, Heard SE. 2006 Annual report of the American Association of Poison Control Centers National Poison Data System (NPDS). Clin Toxicol 2007;45(8):815-917.
3. Hoffman R. Understanding the limitations of retrospective analyses of poison center data. Clin Toxicol (Phila) 2007;45(8):943-5. Editorial.