Pharmaceutical PharmaceuticalSociety

Society ofJapan of Japan

YAKUaAKU

ZASSHI

128(t) 165-170

(2008) @ 2008 The PharmaceuticalSocietyof Japan

165

-Notes-

Treatment for Carbamazepine Detoxication

and

LithiumOyerdose

Hiroko UNEI," HiroakiIKEDA,'･"Teruo MURAKAMI,b KoichiTANIGAwA,C

and

KenjiKIHIRAa

Services,Hiroshima UhiversityHospital,1-2-3 Kasumi, Minami-ku, gDqpartment ofPharmaceutical Hiroshima B4-855J, mpan,bLaboratoi:y and Pharmacokineties,hacully of ofBiopharmaceutics Pharmaceutical Sciences,Hlr'roshimainternationalUniversity,5-1-1 Hiro-koshingai,Kure City rs7-Oi12, Japan and and Critical ofAdvancedEmergenqy Hbspital,i-2-3 Ktisumi, CkereCenter, HiroshimaU}iiversity Minami-ku, Hiroshima Z34-855J, Japan `Dqparttnent

August (Received

9, 2007; AcceptedOctober22,2007;Published online

October23,2007)

This article reports detoxicationtreatments of a case of combined overdose of carbamazepine and lithium ina 38unit after the family found her unresponsive disorder. Shewas brought to the emergency year-oldfemalewith bipolar and lyingnear empty to 7,7 g) and lithium carbonate to 6,6 (corresponded (corresponded packages forcarbamazepine On adrnission, herbloodpressure,heart rate and respiratory rate were 8015S mmHg, 90 per minute and 13 g) tablets. Ml, V1) .She received gastric lavage after intratracheal intubation,folper minute, respectively. Her GCS was 3 (El, lowedby administration of activated charcoal via gastrictube, and a largevolume (800ml/h) of lactateRinger'ssolution by intravenousinfusion.The serum levels of carbamazepine and lithium approximately 5 h after ingestion were 56,O was mainly h charcoal hemoperfusion overdose treateci by a 3 ug!ml and 3.56mEqll, respectively. The carbamazepine levelsby approximately 30-40% as compared with the Iev(CHP)The CHP treatment decreasedserum carbamazepine detoxicationtreatment) .For lithiels simulated by Bayesian analysis using 1-pointor 2-pointsserum level(s)(without um overdose continuous infusionof Ringer'ssolution was effective, which increasedserum sodiurn gradually and facilitatedthe elimination of lithium. In conclusion, the treatrnents with CHP and continuous infusion of Ringer's solution were of carbamazepine and Iithiumoverdose, respectively, when compared considered to be effective for detoxification with those druglevels without detoxication treatment that simulated by Bayesian analysis method, .

Keywords

overdose poisoning; carbamazepine;

lithium; detoxication treatment;charcoal hemoperfusion; Bayesian

analysis

acute,

INTRODUCTION Drug

overdoses

leadto a

variety of

acute

administration

poisoningsym-

on

chronic,

rnodality.

and

The

chronic,

severity

depending on of lithiumin-

toxication isquitedifierent forthese distinct types of

impaired consciousness including

lithium poisoning.i･S) Charcoal hemoperfusion and hemodynamic instability, In such patients, infor(CHP)and injectionsof activated charcoal intothe stomach coupled with aggressive intestinal mation on time profiles of drug levelsin the blood purging would for predicting are clinically effective detoxification methods for carbe highly beneficial (serum) time-dependentpoisoningsymptoms and facilitating bamazepine poisoning.4･9'i3)For lithium poisoning, careful attention must be given to the sodium and timely interventions that prevent adverse efiects,i･2) The presentstudy reports a case of carbamazepine fluidbalance,in addition to promoting renal clearhaye renal and lithium overdose in a 38-year-oldfemalewith ance or hemodialysisfor patientswho disorder. impairment,i,8,i4) bipolar In the present study, the patient with carbamazeCarbamazepine is an anticonvulsant, moeddrug used in the treatment of stabilizing primarily pineand lithiumoverdose was treated with CHP and infusionof Ringer's solution. Also, the epilepsy and bipolardisorders. Lithium is a moodcontinuous disclinical eMcacy of the treatrnents was evaluated by stabilizing drugused inthe treatmentof bipolar orders. Carbamazepine is considered safe when used measuring serum carbamazepine and lithium Ievels with time and by comparing the values with those appropriately, but overdoses can be lifethreatening,3･4'7) Lithium overdoses are dividedinto3 types: drug levels without detoxication treatment that simulatedby Bayesiananalysis method. ptoms

'e-mail:

over

time,

[email protected]

NII-Electronic Library Service

Pharmaceutical PharmaceuticalSociety

Society of of Japan Japan

166

Vol. 128

A CASE PRESENTATION Treatment A 38-year-oldfemale weighing 5S kg with bipolardisorderhad been treated with carbamazepine (tablets, TegretolO,Novartis Pharma K.K, Tokyo, Japan) and lithiumcarbonate (tablets, Limase, Taisho Pharmaceutical Co., Ltd, Tokyo, Japan) for more than two months at the outpatient clinic of Hiroshima University Hospital.During this and lithium period, thedailydoses of carbamazepine carbonate were 700 mg and 600 mg, respectively. One day, the family found her lyingunresponsive near empty packagesof carbamazepine to (corresponding 7.7g) and lithiumcarbonate (6.6 g) tablets.She was transported by ambulance to the emergency room at Hiroshima UniversityHospital after the ingestion. Physicalexarnination on admission revealed; GCS EI. Vl. Ml., bloodpressure at 80155mmHg, heart rate at 90 per m, and respiratory rate at 13 per m. Neither rhythm in disturbance the electrocardiograph nor convulsive seizure was observed, The patientunderwent repeated gastriclavagewith 7000 ml of normal saline followingintra-tracheal intubation,She then received 20 g activated charcoal and 250 rn1 of magnesium liquid, Magcolole, citrate solution (13.6% Horii PharmaceuticalInd.,Ltd, Osaka, Japan)via gastric tube, Subsequently,she received a largeintravenous infusion(800 ml/h) of lactate Ringer's solution(lactate, 28 mEq/1; sodium, 130 mEq/I) .Duringthesetreatments,bloodwas sampled to measure serum carbamazepine and lithium levels. The first sampling time of blood was assumed to be approximately 3 h after the ingestion, since the familywas keepingconsiderable attention to her on the day, The carbamazepine

lithium,as

overdose compared

carbamazepine

was more

severe

than that of

to their therqpeutic Ievels. The

intoxication was treated with

sodium

heparin-added charcoal hemoperfusion (CHP)for 3 h, which

initiated approximately 2 h after admission (5h after ingestion) . Blood was taken 7 times during these treatments to measure carbamazepine, lithiumand sodium levelsin serum. Blood (3ml was each) centrifuged (KubotaKN-70, Kubota Cooperation,Tokyo, Japan) at 10000 rpm for 15 m to obtain serum samples. The serum levelsof carbamazepine were measured by Fluorescence Polarization Immunoassay (FPIA)using the TDX FLX system (Abbott Japan DiagnosticsDivision,Tokyo, Japan), and levelsof Iithiumand sodium were by was

{2008)

Na/K/Li Analyzer (Ciba-Corning 654 Analyzer, Bayer Diagnostics, Tokyo, Japan). Bayesiananalysis was applied to simulate the time course of both drug levelsjn serumi5,i6) on VCM-TDM E edition Ver. 2.04 (Shionogi & Co., Ltd.). After intensive treatment for 3 days, the patient was transferred to a psychiatricward, The permission of

to the the presentation to the scientificjournal

scientific

and

meeting

sub-

of article was obtained

mitting

from the patientafter thecompletion

of

detoxication

treatment.17,r8)

TDM

Data

The

levelsof

serum

carbamazepine

andlithiumimmediatelyafteradmissionwere5O.5ug!

3.20mEqfl, respectively, and at 2 h later, they 56.0"g/ml and 3.56mEq/1. As described,the first sampling tirneof blood was assumed to be 3 h after ingestionof drugs.Using these 1-point(3h) and 2-points(3and 5h after ingestion)serum data for each drug,the serum level-time were obtained profiles analysis method. The simulated by the Bayesian curves using 1-pointand 2-pointsserum data of both carbamazepine and lithiumwere similar as shown in Figs,1 and 2. No noticeable difference in estimated parametersemerged between the l-pointand 2-points 1) in which the absorption Bayesiananalyses (Table rate alone of each drug was difierent from constant ml

and

were

,

the value

reported

in population

pharmacokinetics

studies.]5)

level(56.0uglml) at 5 h after ingestion was approximately 4-foldhigherthan therapeutic level(15 the maxirnal "g/ml) .Thus, the to be overdose of carbamazepine was considered more carsevere than that of lithium,The serum bamazepine levelafter 3-h CHP treatment was lower The

serum

carbamazepine

than the simulated method

CHP

levelby

the

Bayesiananalysis

detoxification treatment) (Fig. 1), (without treatment could

concentration-time

decreasethe

curve

of

serum

area

under

the

carbamazepine

from O to 44 h (AUCe.44) after ingestion, which was calculated by a trapezoidal rule, by approximately 30% of the simulated serum levelsby the Bayesian analysis method detoxification treatment) (without it 2) ,Without the CHP treatment, may take (Table 32 h for serum carbamazepine levelto fallto the upper therapeutic levelin the patient.However, the CHP treatment could reduce the time to 21 h after in1, Table 2). gestion (Fig. mEq!I) was apLithium levelin the patient(3.56 3-fold higher than the maximal therapeuproximately

NII-Electronic NII-Electronic

Library Service

Pharmaceutical PharmaceuticalSociety

Society of of Japan Japan

Ne. 1

167

Table 1. PharmacokineticParametersof Carbamazepine and Lithium in Overdose in a 38-year-oldFemale that Calculatedby the Bayesian Method

Carbamazepine Parameter Elimination Halflife

rate

Population

constant

(h-i)

(h-T) (1!kg)

rate constant

Distribution

vol.

Clearance (lfkg/h) of

1-point2-points

Population

O,e29223,8

O,029223.7

O.029S23.5

1.230

O,354

1.61

1,70

O.047

o.oso

1-point

2-points

e.026626,1

O.026526.2

O,02732S.4

O,389

1.500

O.083

O.070

1,84

e.79

O.79

O.80

O.054

O.021

O.021

O.022

(h)

Absorption

Deses

Lithium

and lithiumcarbonate were assumed to be 7700 mg and 6600 mg, respectively. Fer pharmacekinetic a"alysis, the lagtime befereabO.Populationmeans theaverage value obtained by populationpharmacokinetics data.1-point: Basiananalysis was made using 1-pointserum drug (3h after ingestion) .2-points:Basian analysis was made using 2-pointsserum levelsof each drug (3and 5 h after ingestion) ,

carbamazepine

sorption was set at

levelof

each

Table 2. Effectof Detoxification Treatmentson PharmaeokineticParametersof Carbamazepine and Lithium in

4:

3,5tt 3tt"e"

t 2.5g-..'S

?

I･-

za

2gVL5881Ee

E-1lso

;t g e136 Z

o,ses

oo1224

36

48

8

6072

Time,hrs

Fig,2, Time Courses for Serum Lithium (solid circles) and Sodium Ions (triangles) Following Ingestion of 6.6g Lithium Carbonate in a 38-year-oldFemale A 3-h charcoal hemoperfusion (CHP) was carried out 5 h after ingestion,Closed

'Twodottedlinesrepresent

the time

courses

dafterdetoxicationtreatments. forsemJm carbarnazepine simu-

using 1-point (-) arid 2-points (e) serurn by the Bayesian method levels detoxication treatment) ,---:Therprior to CHP treatrnent (without apeutic range of carbarnazepine is 4-12ptg!ml in serum. Charcoal

lated

Hemoperfusion

di:

(3h>

tic Ievel(1.25 mEq/1)

, During the

constant

elimination

infusionof of

lithium

from plasma appeared to be facilitated, in association with the increasein serum Na concentrations (Fig. 2) .The AUCo-44 of serum lithiumwas lower by 38% than that simulated by the Bayesiananalysis method

with

ly, before,during and represent

broken 1ineand

closed

triangles

with

line

solid

detoxieationtreatments. Two dotted lines for serum lithiumsimulated by the Bayesiarr 1-point(-) and 2-points(e) serum levelspriorto CHP

the tirne

after

courses

method using treatrnent(withoutdetoxicationtreatment).----:Therapeuticrangeoflithium

Iactate Ringer's solution,

circles

representtheobservedserumlithiumandsodiumconcentrations,respective-

isO.4-1.2mEqfmHn

serurn.

er:

Charcoal Hemoperfusion

(withoutdetoxificationtreatment) Without

the

treatment, it may

take 48h

(3h)

(Table2). for

serum

lithiumlevelto fallto the upper therapeutic levelin the patient. The intravenous infusionof lactateRinsolution could reduce the time to 21 h after inger's 2, Table 2). gestion (Fig,

NII-Electronic Library Service

Pharmaceutical PharmaceuticalSociety

Society of of Japan Japan

168

Vol. 128

um

DISCUSSION Accidental or intentional poisoning by drug over-

dose occurs

highrate. Among antiepileptic drugs, carbamazepine isreportedly the second most common overdose drug in U.S,, following valproic acid.i9'22) Four distinctstages of carbamazepine poisoninghavebeenreported. Based on serum carbamazepine levelsthe stages are: potentially catastrophic

at a

at

relapse

ataxia 25 "g/m13･4'7･23'2S)

5 h after (possibly

at

ingestion)

1). Similarly,the severity of (Fig.

Iithiumpoisoning reportedly

correlates

with

serum

Iithium leyels, although

lithium with poisoning differs drug historyas describedabove. Symptoms of severe toxicity, which can be Iife-threatening, include marked delirium,coma and seizures at >2.5mmol/1 in serum, though morbidity and mortality are rare.i･8･26) In the present case, the serum lithium level possibly5 h after ingestionwas 3,S6mEq/I. Carbamazepine is mostly eliminated by hepatic metabolism, with only 1% to 3% renal excretion, and lithiumiselirninated mostly to the urine by glomeruIar fiItrationand secretion.

In the presentcase, the carbamazepine and lithium overdose was mainly treated by gastric lavage,administration of activated charcoal, intravenous volume Ioading,and CHP, Administrationof activatfollowed by the gastric lavage treatment are thought to have effectively removed the drugs ed

charcoal

in the stomach, since the absorption rate for both carbamazepine and lithium estimated by Bayesian analysis were fairlylower than reported values.i5!i6) Bayesiananalysis isperformed by using fractional individual sampatient data (1-4 remaining constants

ples) under netic

steady

parameters

dosage

population pharmacokiis applied to settle adequate

state and and

drugs from individualpharmacokinetic parameters.This techniquehas proven useful for TDM for drugs with narrow therapeutic ranges, includingthe aminoglycosides, digoxin,anticonvulsants, lithiumand theophylline, particularly where drug concentrations are measured during relativelycomplicated dosageregirnens.27) In the present study, the Bayesian analysis method was utilized to simulate the time profiles of carbamazepine and lithischedules

of

overdosed

(2008)

by using 1-pointand 2-pointsserum

drug data and reported populationpharmacokinetic only parameters. As a result, the simulation was made by changing absorption rate constants from reported 1). population pharmacokinetic parameters (Table The carbamazepine was treated with CHP, since the eMcacy of CHP for carbamazepine poisoning has beenwell docurnented.6･9-i3) As shown inFig, 1,CHP increasedcarbamazepine elimination dramatically. In overdose parallel,the lithium travenous sodium and volume absorption

treated

in-

with

Ioading,since the

re-

of lithiumions at the renal proximal tub-

be suppressed

ule could

was

by a higherconcentration

of

ions and a greatervolume of renal-tubule renal excretion of lithium fluid, resulting in increased ions.S)The continuous infusionof Ringer'ssolution sodium

increased serum

gradually, and serum

sodium

lithium

levelsclearly decreasedas a result (Fig, 2) , Hemodialysisis also reportedly a useful detoxification method forlithium,especially in patientswith serum lithium levels higher than 3.5mEqfl or renal

impairmenti,8,28). For patients who ingestmultiple substances simultaneously, itisimportantto set treatment priorities as soon

as

possible. The

severity

of

poisoning

will

vary,

so the detoxification method will differ in each case, In this study, CHP was effective in decreasingserum carbamazepine, but not lithium.Thereforemonitoring and predictingdrug level-time were improfiles in deciding tools treatment portant priorities. The Bayesian dosing program is clinically valid when

two

feedback

levels are

known.29,30) In the

1-pointserum data at 3 h and 2-pointsat 3 and 5h after ingestion. In clinical situations, the 1-point Bayesian analysis may be more useful because there may not be time to take two bloodsamples beforeinitiating treatment, Previously,the Bayesianmethod was applied to a theophylline present study,

overdose

we

compared

(2400mg,sustained-release

tablet)

in a

The initial theophylobserved serum young female.3i) linlevelat 2 h was within the therapeutic Ievel(1O-20

thatthe ug/rd),However, Bayesiananalysis indicated maximal toxic level(32ug/ml)would be reached at 12h after ingestion. Itwas also estimated more than 28 h would elapse beforetheserum theophyllinlevel to the upper therapeutic level.Although decreased only the 1-pointserum theophyllin level (2h after inused for the Bayesian analysis, the gestion) was predictedvalues

were

quite close to the

NII-Electronic NII-Electronic

observed

Library Service

Pharmaceutical Society of Japan

NII-Electronic Library Service

Pharmaceutical PharmaceuticalSociety

Society of of Japan Japan

170

Vol. 128

24) BridgeT.A.,NortonR.L.,RobertsonW.O., Pediatr.Emerg. Care, 10, 260-263 (1994), 25) Stremski E. S., Brady W. B,, Prasad K., Hennes H. A., Ann. Etnerg.A(fed., 25, 624-

630 (1995). 26) BaileyB.,McGuigan M., T7ier. Drug Mbnit., 22, 650-655(2000). 27) Thomson A. H., Whiting B., Clin. Pharmacokinet,, 22, 447-467 (1992) , 28) Kasahara H,, ShinozakiT,, NukariyaK., Nishimura H,, Nakano H., Nakagawa T., Ushijima S.,

.ipn,

-12

(2008)

(1994).

29) GarciaM. J.,AIonsoA. C., MazaA., Santos D,, Matesanz C., Dominguez-Gil A., X CIin, Pharm. Tlher., 13, 375-380 (1988). 30) Gaulier J.M., BoulieuR.,FischerC,, MauF., guiere X Pharm. Pharmacol.,49,734-736

(1997), 31) Yamanoue T,, Okabayashi K., Ikeda H., Iwasaki Y., Kitaura T., Unei H., Inoue T., Wada S.,KihiraK.,OhtaniM,, J. .ipn.Soc, Emer. Med., 3, 438-441 (2000).

X Rsychiatry ?VeuroL, 48, 1

NII-Electronic NII-Electronic

Library Service