Phenytoin intoxication in critically ill patients

CASE REPORTSPhenytoin Intoxication in Critically Ill Patients Gert De Schoenmakere,MD,Jan De Waele,MD,Wim Terryn,MD,Mieke Deweweire,MD, Alain Verstraete,PhD,Eric Hoste,MD,Sylvie Rottey,MD,Norbert Lameire,PhD,and Francis Colardyn,MD●Phenytoin intoxication can result in major and possibly life-threatening disorders.Furthermore,the hepatic clearance can become saturated,thus,shifting the elimination fromfirst-to zero-order kinetics.This results in a slow elimination of the compound in case of intoxication.The treatment modalities for phenytoin overdose are limited.Taking into account the high level of protein binding,the molecule is not easily eliminated from the body by means of extracorporal epuration.Although reports exist on the use of MARS(molecular adsorbents recirculating system)dialysis,peritoneal dialysis,and standard dialysis for the elimination,in practice,hemoperfusion,is the most often applied technique.The authors report the case of a hypoalbuminemic patient with severe neurologic signs of phenytoin intoxication(total concentration moderately elevated,free fraction high).A combination of high-flux dialysis and hemoperfusion resulted in a considerable extraction of the drug,accelerating the natural clearance from the body and ameliorating clinical signs of intoxication.In selected patients(with a high free fraction of phenytoin),high-flux dialysis may be a valuable alternative or adjuvant to hemoperfusion.Am J Kidney Dis45: 189–192.©2004by the National Kidney Foundation,Inc.INDEX WORDS:Phenytoin intoxication;hypoalbuminemia;dialysis;hemofiltration.P HENYTOIN(diphenylhydantoin)is a hydan-toin derivate.It is used to control general-ized tonic clonic and complex partial seizures and for the prevention and treatment of seizures occurring during or after neurosurgery.1In inject-able form it can be used in the treatment of cardiac arrhythmia,especially digitalis-induced arrhythmia whenfirst-line therapy is not effec-tive.2The reported distribution volume of phenytoin is0.5to0.8L/kg body weight.For patients with normal plasma albumin concentrations and in absence of displacing agents,its free fraction is about10%.3,4Phenytoin is metabolized mainly through parahydroxylation by the cytochrome P450system.3,4Eighty percent of the total admin-istered dose of phenytoin appears in the urine as this parahydroxylation product(HPPH).Be-cause of gene polymorphism,a large interpatient variability in phenytoin metabolization has been described.5In the normal dosage range,the elimi-nation is subject tofirst-order kinetics,but high-normal and overdose is characterized by zero-order kinetics because of saturation of the metabolization system.6When marked clinical toxicity occurs,phenytoin elimination is thus no longer dose dependent,rendering the spontane-ous removal of the compound slow and insuffi-cient.Toxicity is characterized by a continuum of neurologic signs and symptoms:nystagmus, ataxia,drowsiness,dysarthria,hyperreflexia, coma,and respiratory depression.Paradoxical seizures have been reported in severe phenytoin intoxication but are extremely rare.7,8 However,in critically ill patients with moder-ate to severe hypoalbuminemia,it has been shown that phenytoin toxicity occurs even when normal plasma levels are obtained owing to an increase in the free fraction.9-11Furthermore,several other predilections of altered phenytoin metabolism coexist in these patients(eg,liver and kidney dysfunction,administration of competitive inhibi-tors).We describe the case of a comatose critically ill patient with an apparently therapeutic total phenytoin level but with a highly toxic free fraction owing to severe hypoalbuminemia.She regained consciousness after3sessions of high-flux dialysis and hemoperfusion.We compare From the Departments of Intensive Care,Nephrology,Neu-rology,and Toxicology,University Hospital,Gent,Belgium. Received April5,2004;accepted in revised form Septem-ber16,2004.Address reprint requests to Gert De Schoenmakere, MD,De Pintelaan185,9000Gent,Belgium.E-mail: Gert.Deschoenmakere@skynet.be©2004by the National Kidney Foundation,Inc.0272-6386/04/4501-0021$30.00/0doi:10.1053/j.ajkd.2004.09.016American Journal of Kidney Diseases,Vol45,No1(January),2005:pp189–192189the efficacy of standard high-flux dialysis with that of hemoperfusion in the treatment of phenyt-oin overdose.CASE REPORTA 49-year-old woman with a history of chronic heart failure,pulmonary embolism,depression,and gastric band-ing operation was transferred to the intensive care unit with acute on chronic liver failure.Three months before,she underwent a Scopinaro opera-tion (gastrojejunostomy ϩresection of the distal two thirds of the stomach).This was complicated by an anastomotic perforation leading to eventration and fecal peritonitis.Two weeks before admission in our unit,phenytoin was started because of partial facial epileptic insults.One week before admission,she had signs of hepatic failure with marked hyperammonemia.At this time,she gradually fell into a coma,with varying degrees of light responsive mydriasis.Electroencephalogram (EEG)results were suggestive of metabolic puted tomography scan of the brain was normal.On admission,she had a hemodynamic profile compatible with sepsis.She was intubated and ventilated but not se-dated.She had marked alterations in the liver function with a blood ammonium level of 121␮g/dL (71␮mol/L)with a normal range from 19to 82␮g/dL (11to 48␮mol/L),and a prominent rise in transaminases with aspartate amino trans-ferase 8times and alanine amino transferase 7times the normal values,respectively.Total bilirubin level was 9.9mg/dL (169␮mol/L),predominantly direct fraction.The phenytoin level was 12.8␮g/mL (51␮mol/L;therapeutic range,10to 20␮g/mL [40to 80␮mol/L]).There was a marked hypoalbuminemia of 1.18g/dL (11.8g/L).Renal function was normal.Because of the abdominal sepsis she was treated with fluconazole (400mg/d)and piperacillin/tazobactam (4ϫ4g/d).She had a Glasgow Coma Scale value of 4of 15.There was no neck stiffness,but a bilateral light responsive mydria-sis was observed.The EEG was compatible with metabolic encephalopathy,the transcranial duplex was negative,and magnetic resonance imaging of the brain was normal.Because of recurrence of the facial epilepsy,the mean phenytoin levels were kept high (20.7Ϯ4.1␮g/mL [82Ϯ16␮mol/L]),and intravenous clonazepam was initiated.Because of the concomitant severe hypoalbuminemia and because of the persistence of coma,we calculated an albumin-corrected phenytoin level (corrected phenytoin ϭmeasured phenytoin/[(albumin *0.2)ϩ0.1]).12For a measured level of 19␮g/mL (75␮mol/L),the true level was 47␮g/mL (186␮mol/L).This was confirmed by direct measurement of free phenytoin:4.7␮g/mL (19␮mol/L;normal values,1to 2␮g/mL [4to 8␮mol/L]).Extracorporal removal of the drug was performed during 3consecutive days for 4hours,taking into account the possible rebound effect after each epuration.For the extracorporal removal,we mounted a high-flux dialyzer (first in line;Fresenius FX80,Fresenius Medical Care,Bad Homburg,Germany)and a charcoal filter (second in line,charcoal cartridge Adsorba 300C,Gambro,Stockholm,Swe-den,contents,300g active carbon)in series.Standard heparin was used for anticoagulation.Blood flow and dialysate flows were 200mL/min and 700mL/min,respectively.To quantify the efficacy obtained by the high-flux dialyzer and the charcoal filter separately,we determined phenytoin concentrations before the dialyzer,between the dialyzer and the charcoal filter,and after the charcoal filter (Fig 1).We found that the average extraction by high-flux dialysis (38Ϯ5%)is markedly superior to that by charcoal (16Ϯ6%).No significant rebound effect was noted.By day 3our patient had regained consciousness and made efforts to speak.The bilateral mydriasis had disap-peared.Over these 3days,no significant changes were observed in liver function (ammonium levels,prothrombin time,liver enzymes,bilirubin,and albumin),no dose adjust-ments were made for clonazepam,and no new sedatives were added or removed.The patient died after 3weeks because of bowel infarction with refractory lactate acidosis andshock.Fig 1.Total phenytoin concentration for 3consecu-tive dialysis/hemoperfusion sessions.At the end of each epuration session,blood samples were taken before the high-flux membrane,be-tween the charcoal filter and the dialysis membrane,and after the charcoal filter.Per-centages demonstrate ex-traction ([concentration be-fore ؊concentration after]/concentration before)over the epuration device.To con-vert phenytoin levels in ␮g/mL to ␮mol/L,multiply by 3.96.DE SCHOENMAKERE ET AL190DISCUSSIONPhenytoin intoxication is a deceptive entity because of the saturation of the hepatic metabo-lization and switch fromfirst-to zero-order kinet-ics at high therapeutic and toxic levels.6This implies that only minor increases in dosage can cause toxic levels.Hence,severe symptoms of toxicity should be treated promptly by extracor-poral elimination techniques.Charcoal hemoperfusion is characteristically used in intoxication with drugs with a relatively small distribution volume,a molecular weight inferior to10,000daltons,and a high plasma protein binding.Some case reports have been published illustrating the efficacy of this epura-tion technique for phenytoin.13-15In1patient13 the expected half-life of24to230hours was reduced to3.9hours by hemoperfusion.It can be hypothesized that the small distribution volume and the small binding constant(Kϭ 6.103/ mol/L)make a substantial elimination by char-coal hemoperfusion possible.A valuable but expensive alternative to he-moperfusion is the use of MARS(molecular adsorbents recirculating system).With this sys-tem,phenytoin is cleared from the blood into the albumin containing dialysate and subsequently cleared from this dialysate using a charcoalfilter. The concomitant use of an anion-exchange col-umn did not alter the phenytoin clearance.A significant reduction,very much superior to the endogenous clearance,could be obtained with significant improvement of the clinical symp-toms.16The value of peritoneal dialysis,however,ad-vocated by some reports as an alternative treat-ment,remains uncertain.17Because of its high protein binding(and hence the low free fraction),standard dialysis seems inadequate for the removal of phenytoin.How-ever,it has been reported that chronic hemodialy-sis patients treated with phenytoin have a higher risk of seizures during or immediately after dialy-sis.18The authors point to the relative increase in free fraction of phenytoin owing to competitive binding of accumulated uremic toxins to serum proteins.This high free fraction leads to more drug removal by standard hemodialysis.In critically ill patients with low serum albu-min levels,the free fraction of phenytoin can be markedly increased compared with standard pa-tients with the same total drug level.9-11Hence,in analogy with phenytoin removal in chronichemodialysis patients,we hypothesized that incritically ill patients with low serum albumin itwould be possible to clear phenytoin signifi-cantly by high-flux dialysis.Given the mean predicted distribution volume of0.65L/kg,3,4the distribution volume in our110-kgpatient would be71.5L.Because the total phenyt-oin concentration before thefirst epuration episodewas15.2␮g/mL(60␮mol/L),the total body con-tent of phenytoin is estimated at1,086mg(4,290␮mol).After correction for the rebound effect after dialysis,the total amount of phenytoin eliminatedduring thefirst and second epuration session(high-flux dialysis and charcoalfilter combined)was350mg(32%of the body content at session start)and171mg(23%of the body content at session start),respectively.Although the reported mean half-life time ofphenytoin in the literature is22hours,inintoxication this may be much longer.In1study19considering the spontaneous elimina-tion of phenytoin in cases of intoxication,theaverage half-life time was about5days(withsome patients having half-life times of morethan7days).We obtained this result with thecombined extracorporeal epuration techniquein3dialysis sessions of4hours,allowingequilibration between body compartments be-tween epuration sessions.In our hemodialysis/hemoperfusion configura-tion we eliminated a considerable amount of thefree phenytoin fraction through the high-fluxmembrane.Consequently,the free phenytoin frac-tion reaching the charcoalfilter was undoubtedlylower than it would have been when charcoalhemoperfusion was carried out without simulta-neous high-flux dialysis.Hence,the calculatedextraction over the charcoalfilter may be under-estimated because our calculations were con-ducted with total phenytoin concentrations.How-ever,during the second epuration session wemeasured free phenytoin before the high-fluxmembrane,between the high-flux membrane andthe charcoalfilter,and after the charcoalfilter;the extraction of free phenytoin was43%overthe high-flux membrane and27%over the char-coalfilter.Furthermore,charcoal hemoperfusionefficacy should be preferentially monitored us-HIGH-FLUX DIALYSIS FOR PHENYTOIN INTOXICATION191ing total drug concentrations,because free as well as bound drug fractions are eliminated via this technique.We found that the combination of high-flux dialysis and charcoal hemoperfusion resulted in a considerable extraction of the drug,accel-erating the natural clearance from the body and ameliorating clinical signs of intoxication. High-flux dialysis is readily accessible and costs less than hemoperfusion.Hence,it could be considered an elegant and efficacious alter-native for or adjuvant to hemoperfusion in case of severe phenytoin intoxication,particularly in the specified settings of renal failure and/or severe hypoalbuminemia.However,additional studies are needed to validate the efficacy of this epuration technique in a larger patient group.REFERENCES1.Asconape JJ:Some common issues in the use of antiepileptic drugs.Semin Neurol22:27-39,20022.Somberg JC,Miura D,Keefe DL:The treatment of ventricular rhythm disturbances.Am Heart J111:1162-1176, 19863.Gugler R,Manion CV,Azarnoff DL:Phenytoin:Phar-macokinetics and bioavailability.Clin Pharmacol Ther19: 135-142,19764.Hvidberg EF,Dam M:Clinical pharmacokinetics of anticonvulsants.Clin Pharmacokinet1:161-188,19765.Schwarz UI:Clinical relevance of genetic polymor-phisms in the human CYP2C9gene.Eur J Clin Invest 33:23-30,2003(suppl2)6.Chua HC,Venketasubramanian N,Tjia H,et al:Elimi-nation of phenytoin in toxic overdose.Clin Neurol Neuro-surg102:6-8,20007.Booker HE,Darcey B:Serum concentrations of free diphenylhydantoin and their relationship to clinical intoxica-tion.Epilepsia14:177-184,19738.Liponi DF:Phenytoin therapy and toxicities.Ann Intern Med101:568-569,19849.Lindow J,Wijdicks EF:Phenytoin toxicity associated with hypoalbuminemia in critically ill patients.Chest105: 602-604,199410.Burt M,Anderson DC,Kloss J,et al:Evidence-based implementation of free phenytoin therapeutic drug monitor-ing.Clin Chem46:1132-1135,200011.Driscoll DF,McMahon M,Blackburn GL,Bistrian BR:Phenytoin toxicity in a critically ill,hypoalbumin-emic patient with normal serum drug concentrations.Crit Care Med16:1248-1249,198812.Anderson GD,Pak C,Doane KW,et al:Reversed winter-tozer equation for normalized phenytoin concentra-tions in trauma and elderly patients with hypoalbuminemia. Ann Pharmacother31:279-284,199713.Kawasaki C,Nishi R,Uekihara S,et al:Charcoal hemoperfusion in the treatment of phenytoin overdose.Am J Kidney Dis35:323-326,200014.Pond S,Rosenberg J,Benowitz NL,Takki S:Pharma-cokinetics of haemoperfusion for drug overdose.Clin Phar-macokinet4:329-354,197915.Baehler RW,Work J,Smith W,Dominic JA:Charcoal hemoperfusion in the therapy for methsuximide and phenyt-oin overdose.Arch Intern Med140:1466-1468,198016.Sen S,Ratnaraj N,Davies NA,et al:Treatment of phenytoin toxicity by the molecular adsorbents recirculating system(MARS).Epilepsia44:265-267,200317.Czajka PA,Anderson WH,Christoph RA,Banner W Jr:A pharmacokinetic evaluation of peritoneal dialysis for phenytoin intoxication.J Clin Pharmacol20:565-569,1980 18.Frenchie D,Bastani B:Significant removal of phenytoin during highflux dialysis with cellulose triac-etate dialyzer.Nephrol Dial Transplant13:817-818,1998 19.Chua HC,Venketasubramanian N,Tjia H,Chan SP: Elimination of phenytoin in toxic overdose.Clin Neurol Neurosur102:6-8,2000DE SCHOENMAKERE ET AL192。